WPS4079 Banking Services for Everyone? Barriers to Bank Access and Use around the World Thorsten Beck, Asli Demirguc-Kunt and Maria Soledad Martinez Peria* Abstract Using information from 193 banks in 58 countries, we develop and analyze indicators of physical access, affordability and eligibility barriers to deposit, loan and payment services. We find substantial cross-country variation in barriers to banking and show that in many countries these barriers can potentially exclude a significant share of the population from using banking services. Correlations with bank- and country-level variables show that bank size and the availability of physical infrastructure are the most robust predictors of barriers. Further, we find evidence that in more competitive, open and transparent economies, and in countries with better contractual and informational frameworks, banks impose lower barriers. Finally, though foreign banks themselves seem to charge higher fees than other banks, in foreign dominated banking systems fees are lower and it is easier to open bank accounts and to apply for loans. On the other hand, in systems that are predominantly government-owned, customers pay lower fees but also face greater restrictions in terms of where to apply for loans and how long it takes to have applications processed. These findings have important implications for policy reforms to broaden access. JEL Classification: G2, G21, O16 Keywords: financial development, banking sector outreach, financing obstacles World Bank Policy Research Working Paper 4079, December 2006 The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Policy Research Working Papers are available online at http://econ.worldbank.org. *The authors are with the World Bank's research department. We thank Jerry Caprio, Stijn Claessens, Xavier Giné, Patrick Honohan, Leora Klapper, Inessa Love, and Susana Sánchez for comments and suggestions. Edward Al- Hussainy, Andrew Claster, Subika Farazi, Ning Jiang, and Hamid Rashid provided excellent research assistance. 1. Introduction To open a checking account in Cameroon, you need over 700 dollars, an amount higher than the GDP per capita of that country, while no minimum amounts are required in South Africa or Swaziland. Fees to maintain a checking account exceed 25 percent of GDP per capita in Sierra Leone, while there are no such fees in the Philippines. The fees for transferring 250 dollars internationally are 50 dollars in the Dominican Republic, but only 30 cents in Belgium. While most people in the developed world take access to banking services for granted, price and non-price barriers prevent large parts of the population in developing countries from accessing and using formal banking services. If we follow previous estimates (Genesis, 2005b) that poor people cannot afford to spend more than 2 percent of their household income on bank charges, the fees observed in many countries effectively prevent them from using such accounts. Similarly, the requirement of a physical address or of a formal sector job as eligibility criteria to open an account excludes the majority of people in many developing countries, where a large percentage of the population lives in rural areas and works in the informal sector. This paper presents new indicators of barriers to bank access and use of banking services around the world, shows their significance for outreach and relates them to bank and country characteristics. First, through surveying the largest banks in 58 countries, we document the extent of barriers to three banking services - deposits, loans and payments - across three dimensions - physical access, affordability, and eligibility. Second, we show the importance of these barriers for access to and use of financial services. Third, we explore which bank and country characteristics are associated with these barriers, with findings that have important implications for policies to broaden access. Market frictions such as transaction costs and information asymmetries give rise to financial institutions and markets (see Diamond 1984, 1991, Ramakrishnan and Thakor 1984, 1 Boyd and Prescott 1986). These market frictions, however, can also limit the extent to which financial institutions can reach out to clients. Transaction costs that to a large extent are independent of the size of the financial transaction ­ deposit, loan or payment ­ make outreach to clients with demand for small transactions costly. High information asymmetries and the resulting agency problems make outreach to opaque clients more difficult and costly. Barriers such as high minimum account balances and fees, multiple documentation requirements and high payment fees might reflect high transaction costs and the contractual and business environment in which banks operate. However, they might also reflect the competitive framework and the availability of physical infrastructure in the market where banks offer their services. Financial exclusion can retard economic growth and increase poverty and inequality. Theoretical models have shown that financial market frictions that prevent broad access can be the critical mechanism for generating persistent income inequality or poverty traps (Banerjee and Newman, 1993; Galor and Zeira, 1993). A large empirical literature has established the importance of banking sector depth for economic development and poverty alleviation. Based on extensive cross-country databases, researchers have explored the relation between indicators of financial sector depth and GDP per capita growth, productivity growth, poverty, firm growth and entry rates (see King and Levine, 1993; Beck, Levine and Loayza, 2000; Demirguc-Kunt and Maksimovic, 1998; Rajan and Zingales, 1998; Beck, Demirguc-Kunt and Levine, 2004; Klapper, Laeven and Rajan, 2006). Much less is known, however, about the determinants and implications of access to financial services by individuals and firms. This is because data on who has access to which financial services remain thin and inadequate. This paper contributes to closing this gap in the literature. Our data show substantial cross-bank and cross-country variation in barriers to banking. While banks in 19 countries do not impose any minimum balances for checking accounts, such 2 balances are higher than 10 percent of GDP per capita in 14 countries. While one document is needed to open an account in five countries, four documents, including ID, payment slip, proof of domicile and reference letter, are required in six countries, effectively preventing large parts of the population from accessing these services. While it is possible to apply for a loan over the phone or the Internet in six countries, customers can only submit loan applications at bank headquarters or at branches in five countries. We conduct consistency checks on our data and show that, in general, banks in more economically and financially developed economies impose lower barriers. Also, we show that barriers are negatively correlated with financial outreach ­ measured by branches, loans and deposits per capita ­ and with lower financing obstacles as reported by firms. Our indicators thus capture an important aspect of banking sector outreach across countries. While double-digit ratios of minimum balances, fees and minimum loan amounts to GDP per capita already give a first impression of the limited affordability of many of these services for large parts of the population in a number of countries, we offer back-of-the-envelope calculations using data on income distribution. We find, for example, that fees to maintain checking accounts effectively prevent more than 30 percent of the population from using such services in ten of the 58 countries in our sample. We also explore which bank- and country-level characteristics are associated with barriers to banking. Consistent with the argument of scale economies in banking, we find that larger banks consistently impose lower barriers. Better physical infrastructure is also robustly associated with lower barriers. In more competitive, open and transparent economies bank customers also face lower barriers. Further, we find evidence that banks in countries with a more efficient contractual and informational framework impose lower barriers on customers. Finally, the relationship between barriers and bank ownership is not a simple one. Though foreign banks 3 themselves seem to charge higher fees than other banks, in foreign dominated banking systems, fees are lower and it is easier to open bank accounts and to apply for loans. On the other hand, in systems that are predominantly government-owned, customers pay lower fees but face greater restrictions in terms of where to apply for loans and how long it takes to have applications processed. This paper is related to an emerging literature on access to financial services. Most of the existing research and the efforts underway focus on country case studies that aim at measuring and analyzing access to financial services at the household or firm level (see Claessens, 2006 and Claessens and Demirguc-Kunt, 2006). Few papers study this issue by focusing directly on banking services providers. Recently, Beck, Demirguc-Kunt and Martinez Peria (2007) present aggregate cross-country data on banking sector outreach (such as branch and ATM penetration, deposits per capita, and loans per capita) and show that these indicators closely track more difficult and costly to collect micro-level statistics of household and firm use of banking services. More directly related to our paper, Genesis (2005a) examines the costs of using bank accounts in seven countries - Brazil, India, Kenya, Malaysia, Mexico, Nigeria, and South Africa. However, in contrast to our study, this report focuses exclusively on deposit service affordability in a small number of countries. While our paper is the first systematic effort to document and analyze banking barriers across countries, it has a number of limitations. First, our attempt to compare standard products across a broad sample of countries is limited by differences in financial practices. For example, while in some countries checking accounts are the prevalent form of transaction account, in other countries checking accounts might not be widely used and savings accounts might be preferred. Furthermore, even the same type of financial product, e.g., an SME loan, might have different definitions and features across banks and countries. We therefore assess barriers on somewhat 4 different deposit and loan products. However, to the extent standardized products are not offered across countries, it is difficult to overcome this problem.1 Second, fees and charges might differ because of differences in the scope and quality of the services provided rather than because of differences in pricing strategies. Third, we focus on the largest banks, not on the whole banking system. While this seems a restriction, by focusing on the largest banks we capture the barriers encountered by a majority of customers in a country. Finally, the nature of our survey is such that we are not able to capture non-bank financial institutions, such as finance companies and microfinance institutions. In spite of these shortcomings, we see this paper as an important first step in the effort to create consistent cross-country indicators of barriers that households and firms face in accessing financial services. The remainder of the paper is organized as follows. Section 2 discusses the survey used to collect bank-level information. Section 3 presents the indicators and discusses their cross- country variation. Section 4 shows that these barriers are correlated with cross-country indicators of outreach and firms' financing obstacles and section 5 offers back-of-the-envelope calculations that show the impact of some of these barriers on access. Section 6 relates our indicators to bank and country characteristics associated with the institutional, contractual and, competitive environment, and section 7 concludes. 2. The survey The dataset is constructed from a web-based survey with 75 questions that was sent to the five most important banks in 115 countries in 2004 and 2005.2 We chose to focus on the largest 1We also considered asking questions on standardized loans and deposits, yet decided to collect information on actual barriers as opposed to "hypothetical" ones based on products that might not exist in all countries. 2We defined importance in terms of total assets or branches. Data collected from bank regulators and analyzed by Barth, Caprio, and Levine (2004) indicates that on average the five largest banks in over 100 countries account for 73 percent of bank assets and deposits. 5 banks since we are interested in the barriers encountered by the average customer in each country. Survey responses were carefully "cleaned" through extensive follow-up with the banks whenever we had questions about the data provided. While we received a total of 253 responses from banks in 88 countries, to insure representativeness, we limited the analysis in this paper to countries for which the responding banks constitute at least 30% of the market in terms of total loans/total deposits or where we received a response from the largest bank.3 This gives us a total sample of 193 banks across 58 countries. Table 1 presents all the countries in our sample and shows their level of economic and financial development, as measured by GDP per capita in U.S. dollars and private credit to GDP, respectively. Also, the table contains information on the number of banks (out of the top 5 banks) that responded to our survey, along with the market share that they represent. Our sample comprises countries across all levels of financial and economic development. Countries range from Ethiopia with a GDP per capita close to 100 dollars to Switzerland, where GDP per capita exceeds 34,000 dollars. With banking sector credit at 2 percent of GDP, Mozambique is the country with the lowest level of financial development in our sample, while Denmark and Switzerland rank at the top with private sector credit exceeding 150 percent of GDP. In terms of regions, our sample coverage is also quite balanced. Our dataset includes 15 countries from Eastern Europe and Central Asia, 13 countries from Sub-Saharan Africa, 9 countries in Western Europe, 8 Latin American and Caribbean countries, 5 countries from the Middle East and North Africa, 4 countries in South Asia, 3 countries in East Asia and 1 non-European developed country (Australia). 3We determined the market share using data from Bankscope. We have data for the largest bank constituting less than 30% of the market in only one country, Swaziland. In Algeria too, we only have data for the largest bank, but this bank accounts for more than 30% of the market. 6 In terms of market share, for 56 out of the 58 countries in our sample the share of deposits captured by respondents exceeds 30 percent. Banks from France and Zimbabwe are not included in the calculations for deposit and payment barrier indicators because the market share of bank respondents in these countries is below 30 percent. When it comes to loans, the share represented by bank respondents exceeds 30 percent in 53 countries. In this case, the countries excluded from the sample are Germany, Nigeria, Romania, Swaziland, and Sweden. In 34 (29) countries the share of deposits (loans) exceeds 50%. On average across countries, the banks that responded to our sample account for 57 percent of the deposits and 53 percent of the loans in the countries in our sample, based on data from Bankscope. 3. The indicators This section presents our indicators of barriers to banking across countries. Tables 2, 3 and 4 present the country-level averages including descriptive statistics and Figures 1 through 16 show the cross-country variation graphically. Table 5 reports correlations across the different barriers. We separate our indicators based on the type of service: deposit, loan and payments. We report averages for each country calculated by weighing each banks' responses by their share of deposits in total deposits of all sampled banks in the case of deposit and payment barrier indicators and by the share of loans for loan barrier indicators. Also, wherever possible, we try to distinguish between three different service dimensions: physical access, affordability, and eligibility. Physical access refers to the points of service delivery. Greater physical access means services are delivered in multiple and more convenient ways. Affordability refers to the costs in terms of minimum balances and fees that bank clients need to pay to obtain financial services, such as checking or savings accounts, consumer or SME loans, international payment transfers and use of ATM cards. Finally, eligibility refers to the criteria (in terms of documents or other 7 requirements) that determine who can access financial services and who cannot. Regulatory requirements, e.g. in the context of anti-money laundering legislation, might force banks to impose such eligibility requirements. In the case of lending, we use the days needed to process a loan application as an eligibility criterion since some potential bank customers might not apply for loans if they need financing urgently and they know it takes a long time to get a decision. 3.1. Deposit services The main products we consider in terms of deposit services are the checking and savings accounts. Across countries, there are differences in the extent to which savings or checking accounts are the dominant transaction account type. We therefore assess barriers to deposit services based on survey questions related to both account types. Potential customers can encounter barriers to the use of deposit services in terms of the need to visit headquarters to open an account instead of doing it in the local bank branch or a non-branch office (physical access), payment of high minimum balances and fees (affordability), and the requirement to present multiple documents to open an account (eligibility). We will discuss each of these barriers in turn. Weighted country-level averages are presented in Table 2. Physical access Physical access to banking services can often be hampered by long distances from the next bank outlet (Beck, Demirguc-Kunt and Martinez Peria, 2007). However, even if there is a sufficiently wide network of bank offices, not all of these offices might offer the same services. We measure physical access in deposit services by considering the locations to open a deposit account. This indicator takes values from 1 to 3 depending on whether an account can be opened at headquarters only (1), at headquarters or a branch (2) or at headquarters, branches or a non- 8 branch office (3).4 While the majority of sampled banks in Greece and Sierra Leone require customers to visit the head office to open a checking account, customers in Moldova can open such an account at headquarters, branches and even branch-like offices. Overall, we find a substantial variation in the locations to open a deposit account (Figure 1). In the median country, customers can open accounts at headquarters or branches but not at non-branch offices. Affordability We characterize the affordability of deposit services across countries by looking at the minimum balances required to open checking and savings accounts, along with the fees needed to maintain such accounts. There is substantial variation in the ratio of the minimum balance needed to open a checking account to GDP per capita (Figure 2). While in Cameroon and Nigeria, the minimum balance exceeds 100 % and in Ethiopia, Sierra Leone and Uganda, more than 50% of per capita income is required to open a checking account, the amount is zero in 19 countries, less than half of which are developed.5 The median value for this indicator is 0.6 % and the average is 10.9%. While some of the variation in this indicator might be explained by the denominator ­ GDP per capita ­ the correlation between the amount necessary to open an account and GDP per capita is far from perfect (-0.28) and even in dollar terms, there is a significant variation in minimum balances. The ratio of the minimum balance needed to open a savings account to GDP per capita (Figure 3) ranges from zero in nine countries (i.e., Australia, Belgium, Chile, Denmark, Egypt, Israel, Spain, Switzerland and Turkey) to over 40% in Cameroon, Kenya, Sierra Leone 4We consider only the most local office, i.e. banks that allow customers to open an account at a branch or a non- branch office receive the same rating (3) as banks that allow customers to open an account at headquarters, a branch or a non-branch office. 5Countries for which the minimum balance to open a checking account averages zero include: Algeria, Australia, Belarus, Belgium, Brazil, Croatia, Denmark, Georgia, Germany, Israel, Lithuania, Malawi, Moldova, South Africa, Spain, Swaziland, Sweden, Switzerland, and Turkey. 9 and Uganda. The median value for this indicator is 1.1%. The required minimum balance to open a savings accounts is on average only slightly below the minimum balance in checking accounts, 8.1% for the former compared to 10.9% of GDP per capita for the latter. As reported in Table 2, there is a similar variation across countries in the balances that have to be maintained in checking and savings accounts. Thus, the affordability barriers expand beyond the initial stage of opening a checking or savings account. There is a high correlation between the amounts needed to open and to maintain checking and savings accounts, although on average, the amounts are significantly lower to maintain than to open an account, 2.9% and 6.2% of GDP per capita for checking and savings accounts, respectively.6 Fees associated with maintaining a checking or savings account also vary significantly across countries (Figures 4 and 5). While in Malawi, Uganda and Sierra Leone, checking account fees amount to over 20% of GDP per capita, these accounts are free in Bangladesh, Belarus, Ethiopia, India, Jordan, Malta, Pakistan, Philippines, and Sweden. The median value for the fees associated with checking accounts is 0.3% and the average is 2.4%. Savings accounts fees are significantly lower than those associated with checking accounts, ranging from zero in 29 countries to almost 4% of GDP per capita in Malawi and Uganda. The average value across countries for the fees on savings account is 0.4% while the median is exactly zero. Eligibility Around the world, banks demand proof of identification to open an account for a new client. However, banks in many countries demand a variety of other documents on top of ID cards, including recommendation letters, wage slips, and proof of domicile. To quantify these 6Given the high correlation between minimum balances to open and to maintain accounts, we will focus on the minimum balances to open an account in the subsequent analysis. 10 eligibility requirements, we create indicators of the number of documents required to open checking and savings accounts, respectively. While banks in Albania, Czech Republic, Mozambique, Spain and Sweden demand on average only one document to open a checking account, banks in Bangladesh, Cameroon, Chile, Sierra Leone, Trinidad and Tobago, and Uganda require at least four documents (Figure 6). On average, a slightly smaller number of documents is required to open a savings account (2.1) relative to a checking account (2.5). In 10 out of 52 countries for which information is available on the number of documents needed to open a savings account, only one type of document is required.7 On the other hand, more than three documents are needed in Bangladesh, Cameroon, Ghana, Malta, Sierra Leone, South Africa, and Trinidad and Tobago (Figure 7). 3.2. Credit services We collected indicators of physical access, affordability and eligibility for four different loan types ­ business, SME, consumer, and mortgage loans. However, due to space constraints and because of our interest in products available to individuals and to typically constrained smaller firms, we focus on consumer and SME loans (see Table 3). Nevertheless we report indicators on the other loan types in Appendix Table A.1. Indicators of physical access, affordability and eligibility barriers are highly correlated with each other across the different loan types. Physical access To measure physical access for loans, we examine the locations to submit a loan application. While customers in Algeria, Armenia, Ethiopia, Sierra Leone and Uganda can only 7These countries include: Albania, Algeria, Belarus, Czech Republic, Hungary, Lithuania, Mozambique, Spain, Sri Lanka, and Sweden. 11 apply for loans at a bank's headquarters and branches, customers in Australia, Chile, Denmark, Greece, South Africa and Spain not only can use branch and non-branch outlets, but even submit loan applications over the phone and the Internet (Figure 8). In the median and average country, bank customers can submit loan application at headquarters, branch and branch-like offices. Affordability We measure loan affordability by looking at the minimum balances required for consumer and SME loans and at the fees for these loans. The minimum amount for consumer loans relative to GDP per capita ranges from less than 1 percent in Denmark and Switzerland to 330 percent of GDP per capita in the Philippines (see Figure 9). The median minimum amount for consumer loans is 18.54 percent, while the average is 52.29 percent. While banks in Algeria, Belarus, Denmark, and Egypt do not specify minimum amounts for SME loans, banks in Bangladesh set a minimum of almost 10,000 percent of GDP per capita and those in Uganda and Georgia report a minimum of over 2,000 percent of GDP per capita (Figure 10). These very high minimum loan requirements suggest that in those countries banks do not meet the external financing needs of smaller enterprises. The average minimum amount for SME loans is 558 percent and the median is 58 percent of GDP per capita. Fees on consumer loans expressed as a percentage of minimum loan amounts range from zero in Algeria, Belgium, Ethiopia, and Switzerland to over 20 percent of the minimum loan amount in Chile and the Dominican Republic (Figure 11). The median fee on consumer loans is 2 percent and the average is 4 percent. 12 Fees on SME loans also exhibit a significant cross-country variation. Fees vary from zero in Algeria and Switzerland to close to 30 percent in the Dominican Republic (Figure 12). The average fee on SME loans across countries is 3 percent and the median is 1 percent.8 Eligibility A crucial function of financial intermediaries is to screen borrowers beforehand and to monitor them during the lifetime of a loan. However, the number of days it takes to process a loan application can be perceived as a de facto eligibility barrier, since some borrowers might not apply for bank loans and seek financing elsewhere to avoid long waiting periods. For consumer loans, this indicator ranges from almost one day in Australia, Brazil, Czech Republic, Denmark, Greece, Israel and Spain to over 20 days in Pakistan (see Figure 13). The average number of days to process a loan application is 4 and the median is closer to 3. SME loan application are processed in less than 2 days in Denmark and Spain but take more than one month to process in Bangladesh, Pakistan, and Philippines (Figure 14). Across countries, it takes an average of almost 11 days to process a loan application. The median number of days is 8. 3.3. Payment services Our indicators on payment services measure primarily affordability. We examine the costs of transferring a small amount of funds internationally and the fees associated with using ATM cards (see Table 4).9 8We also computed loan fees relative to GDP per capita, as the ratio of loan fees to minimum loan amounts might also represent variation in minimum loan amounts additional to variation in fees. While that ratio gives different rankings of countries, the results reported in sections 4 and 6 do not differ across ratios of fees to minimum loan amounts or GDP per capita. 9Though ATM cards can be used for transactions other than withdrawing cash (e.g., transferring funds across accounts), we think of ATMs as primarily facilitating payments by allowing the withdrawal of funds. 13 The cost of transferring funds internationally varies from 0.12 percent in Belgium to 20 percent in the Dominican Republic (Figure 15).10 To compute these ratios and to make them comparable across countries, we focus on a typical transfer of 250 dollars. On average, the cost of transferring funds internationally is 6.5 percent or $ 16.35. We express the fees associated with ATM transactions as a percentage of 100 dollars. We find that ATM fees are above 0.4 for Pakistan and Nigeria, average 0.1 across countries while the use of ATM is free for 50 percent of the sample (Figure 16). 3.4. Correlations Table 5 shows the pairwise correlations between the different barrier indicators, averaged on the country level. Most of the variables are significantly correlated with each other, although the correlations are stronger among indicators of the same type of service (deposit, loan or payment) than between indicators across the different services. Among deposit service indicators, we find that banks in countries with high minimum balances for checking accounts also require high minimum balances for savings account, as expected. Also, fees are correlated across account types and higher checking fees are positively correlated with higher minimum checking and savings deposit balances required to open deposit accounts. In countries with high deposit fees and high minimum balances, prospective depositors are also required to present a larger number of documents to open accounts. Loan indicators are also correlated with each other but to a lesser extent than is the case among deposit indicators. SME and consumer loan fees are significantly correlated with each other and so are the days to process SME and consumer loan applications. The indicators on the number of days to process loans are also positively correlated with minimum loan balances. 10While we also considered the speed of transfers in terms of days, we found little variation across banks and countries. 14 Among the payment service indicators, the cost to transfer funds internationally is positively correlated with the fees associated with using ATM cards. Across the three different types of services, we find that countries with higher minimum loan amounts also tend to have higher minimum deposit amounts and, in the case of consumer loans, also higher checking fees. Also, we find a significantly positive correlation between the number of documents required to open accounts and the days to process loan applications. Finally, we observe that higher loan fees are correlated with higher costs of transferring funds internationally and higher fees for using ATM cards are positively associated with more requirements to open deposit accounts. 4. Barriers to banking, financial and economic development, and outreach In this section we explore the association between our barrier indicators and existing measures of financial and economic development, as well as of financial outreach (Table 6). In many ways, examining these correlations represents a consistency check on our indicators. As expected, we find that barriers to banking are negatively correlated with economic development. Specifically, minimum balances to open accounts and fees to maintain them, the number of documents to open accounts, the minimum amount of consumer loans, the days to process consumer and SME loans, and the fees for using ATM cards are negatively and significantly correlated with GDP per capita. In the same way, we find that the number of places to submit loan applications, an indicator of lower barriers to physical loan access, is positively and significantly correlated with GDP per capita. Further, we find that higher barriers are consistently negatively associated with financial development. Table 6 shows that private credit to GDP ­ a standard measure of financial intermediary development ­ is negatively and significantly correlated with the minimum 15 balances to open accounts, the annual fee and the documents to open checking accounts, the minimum amount for consumer loans, the days to process SME and consumer loans, and the fees for using ATM cards. On the other hand, private credit to GDP is positively and significantly correlated with the number of locations to submit loan applications. Interestingly, the fees on consumer and SME loans, the cost to transfer internationally and the locations to open deposit accounts are not significantly correlated with economic or financial development. To gauge the relationship between barriers and aggregate measures of financial sector outreach, we utilize recently compiled data on branch penetration and the number of loan and deposit accounts (Beck, Demirguc-Kunt and Martinez Peria, 2007). These are country-level indicators, compiled from regulatory surveys and publicly available information. We would expect countries with banks that impose higher barriers on their customers to invest less in the number of branches and higher barriers to be also reflected in fewer deposit and loan accounts per capita. The correlations in Table 6 suggest that higher barriers are indeed associated with lower outreach. Specifically, banks in countries with a higher demographic branch penetration demand lower minimum balances to open accounts, require fewer documents to open accounts, are more likely to accept loan applications in branch-like offices or over the phone or Internet, set lower minimum loan amounts, are quicker at processing loan applications, and charge lower fees for using ATM cards. Similarly, banks in countries with higher loans per capita have lower minimum loan amounts, are quicker in processing loan applications and are more likely to accept these applications outside headquarters and through non-traditional channels such as phone or Internet. Banks in countries with more deposits per capita demand lower minimum balances and lower fees, require fewer documents to open such an account, set lower minimum amounts for 16 consumer loans, are faster in processing loans and are more likely to accept loan applications through non-traditional channels. These correlations are simply that ­ correlations. They do not imply causality. They show, however, that our indicators capture an important dimension of financial sector development: the limited outreach of the banking system implied by higher barriers. They suggest that barriers to banking go hand in hand with less physical access to banking offices and lower use of deposit and credit services by households and firms. Finally, higher barriers are associated with higher financing obstacles as reported by firms. We use responses to firm-level survey questions on "Is access to financing (e.g. collateral) a problem to the operation and growth of your enterprise?" and "Is cost of financing (e.g. interest rates) a problem to the operation and growth of your enterprise?" from the Investment Climate Assessment (ICA) surveys conducted by the World Bank across 36 (access) and 37 (cost) countries. Responses to these questions are coded between zero (no obstacle) to four (very severe obstacle), with higher values thus indicating more severe financing constraints.11 We take the average across all firms in a country. We find that firms report higher financing obstacles in countries where banks impose higher minimum amounts to open checking and savings accounts and charge higher fees to maintain these accounts, where banks do not accept loan applications through non-traditional channels and take longer to process SME loan applications. Finally, firms report higher financing obstacles in countries where banks demand a larger number of documents to open bank accounts. It is interesting to note that firms' financing obstacles are more significantly correlated with barriers related to deposit services than with barriers related to payment or loan services. This suggests that firms rely to a large extent not only on credit services, but on a whole array of financial services from financial institutions. 11There is a growing literature that shows the importance of financing obstacles for firm growth and financing patterns (Beck, Demirguc-Kunt and Maksimovic, 2005; Ayyagari, Demirguc-Kunt and Maksimovic, 2006). 17 5. Financial exclusion ­ the effects of banking barriers This section provides back-of-the-envelope calculations of the effects of barriers in terms of the percentage of the population in a country that cannot afford banking services. Specifically, we combine income and income distribution data with our information on annual fees to maintain checking and savings accounts to compute the share of the population that does not earn enough to afford using checking and saving accounts (see methodological explanation in the appendix). Using the latest income distribution data from UNU-WIDER (2005), we utilize information on the Gini coefficient to compute percentiles of income distribution and combine this with income data to compute income per capita data at different percentiles of the income distribution.12 We follow Genesis (2005b) and assume that people cannot afford to spend more than 2% of their annual household income on financial services.13 We adjust income with the average household size for every country.14 These calculations provide us with a cut-off percentile of a country's income distribution below which the use of checking and saving accounts is not affordable. Table 7 shows that while in terms of fees checking and savings accounts are affordable for almost the entire population in many countries, there are significant outliers. In ten countries at least 30% of the population cannot afford checking accounts and in several African countries, more than 50% of the population is priced-out of using these services. Specifically, 54% of the population in Cameroon, 81% in Kenya, 40% in Madagascar, 94% in Malawi, 89% in Sierra 12Calculations are based on Dollar and Kraay (2002) and Lopez and Serven (2006). 13According to Genesis (2005b), the 2% limit is based on unpublished research by the South African Universal Services Agency in the context of mandated rolling-out of telecom service to lower-income families. As both financial transaction accounts and telecom service can be considered network products, similar assumptions on affordability for both services seem reasonable. 14Household size is expected to vary with income level within countries. As we do not have data available on household size distribution, we are not able to adjust for this effect. Again, our numbers are indicative and a more detailed analysis would require richer country-level information on the variation of household size distribution with income distribution. 18 Leone and 93% in Uganda cannot afford the fees for checking accounts, given their annual income and assuming that they cannot spend more than 2% of household income on financial transaction account charges. The fees on savings accounts are in general less restrictive. Approximately, 33% of the population in Malawi and Uganda and 17% of the Bolivian population cannot afford the fees and charges associated with a savings account. While these computations are rough estimates, they are most likely conservative estimates of the share of the population that cannot afford these services, as we do not take into account the costs imposed by minimum balances, restricted locations to access services, and documentation requirements. 6. What explains banking barriers across banks and countries? This section explores which bank and country-level characteristics can explain the wide variation of barriers across countries. Theory suggests the importance of transaction costs, risk mitigation tools and market structure for the cost of financial services and thus barriers to banking. Theory also suggests that different business strategies, the size of a bank and its ownership structure might impact its cost structure and, thus, the barriers imposed on customers. We therefore consider whether the size, business orientation and ownership of the banks are associated with barriers and explore the role of physical infrastructure, costs of doing business, the contractual and informational frameworks, banking sector market structure, competitiveness, openness, and transparency of the economy in explaining cross-country variation in banks' barriers. While bank-level data are from Bankscope, country-level variables are drawn from different databases.15 Appendix Table A.2 shows definitions and sources for the explanatory 15Bank ownership data are from Micco, Panizza and Yañez (2007), based on cleaned Bankscope data. 19 variables included in the analysis and Tables A.3 and A.4 present descriptive statistics and correlations for all explanatory variables. To assess the relationship between barriers and bank- and country-level characteristics, we utilize the following regression model Fi,k =0 + 1 Bi + 2 Ck + i,k , (1) where F is one of the barriers indicators for bank i in country k, B is a matrix of bank-level variables (the log of total assets in U.S. dollars, dummy variables for government and foreign ownership and the loan to asset ratio), and C is a country-level variable. While we include all bank variables in our regressions, we include only one country-level variable at a time given the limited number of countries in our sample and the high correlation between our variables (Appendix Table A.4). Critically, we do not control for GDP per capita because many of our explanatory country-level variables are highly correlated with economic development. Also, we are interested primarily in which components of economic development can explain cross- country variations in barriers, as captured by individual country characteristics. We utilize different estimation techniques depending on the nature of the dependent variable. Specifically, for all affordability indicators ­ constructed as minimum amounts and fees relative to GDP per capita-, we conduct OLS regressions of the log of one plus the variable ­ to account for the skewed distribution of these variables. Similarly, for the days to process loans and documentation requirements to open an account, we use OLS regressions. For the location variables (both for loans and deposits) capturing physical access, we utilize ordered probit estimations to take account of the polychotomous nature of these variables with natural order. In all cases, we drop the top 1% of the distribution of the dependent variables to control for outliers. The first four rows of Table 8 report the results of a regression on just the bank-level 20 variables, while all subsequent rows report the results of adding the country-level variables one at a time. Theory provides opposing views on the impact of size, business orientation and ownership on barriers. On the one hand, large banks might be better in exploiting scale economies, thus overcoming more easily the triple problem of smallness faced by financial systems in large parts of the developing world which have clients with demand for small and few transactions and have few customers over which fixed transaction costs can be spread (Beck and de la Torre, 2007). On the other hand, small banks might be closer to these "smaller" and riskier clients and/or orient themselves more towards them (Berger, Hasan and Klapper, 2004). Banks that are less interested in retail business might impose higher barriers to signal this lack of interest to potential customers. While the public-interest theory justifies the creation of government-owned banks with the necessity to target small and riskier clients ignored by private financial institutions, a large theoretical and empirical literature suggests a mission drift of these banks (La Porta, Lopez-de-Silanes and Shleifer, 2002), with both hypotheses having opposing implications for the barriers imposed by government-owned banks. Finally, while foreign- owned banks are assumed to be more interested in large corporates and private clients with demand for large transactions due to their limited access to soft local information (Mian, 2006), they might have more efficient technologies, which allows them to lower cost and thus barriers (Berger and Udell, 2006). We measure the size of banks with the log of total assets in millions of US dollars, the business orientation with the loan-asset ratio (Laeven and Levine, 2006) and their ownership with separate dummy variables for majority government- and foreign-owned banks. Our results suggest that larger banks demand lower minimum balances to open a checking or savings account, charge lower checking and savings fees, require fewer documents to open accounts, impose lower minimum loan amounts for SME and consumer loans, need 21 fewer days to process loans, and are more likely to accept loan applications through non- traditional channels such as phone or Internet. We find that foreign banks appear to charge higher deposit fees and fees on consumer loans, while government-owned banks take longer to screen loan applications. The correlation between business orientation and barriers is mixed. While, retail, loan intensive banks ­ those with higher ratio of loans to assets - require lower minimum balances to open accounts, they ask for more documents to open accounts and take longer to process loan applications. Overall, these results suggest that size is the dominating bank characteristic in explaining variation in barriers. While the academic literature has paid surprisingly little attention to the relationship between infrastructure, input costs and financial depth and breadth, our results suggest that the quality of physical infrastructure, such as communication and electricity networks, impacts the costs of doing business for banks and can explain cross-country variation in many barriers to banking. We use two indicators to gauge the relationship between physical infrastructure and barriers to banking. Specifically, we utilize telephones lines per capita and electric power transmission and distribution losses as percentage of output (Estache and Goicoechea, 2005). Our regression analysis suggests that banks in countries with better phone networks demand lower minimum amounts to open checking or savings accounts, charge lower account fees, require fewer documents to open accounts, allow loans to be submitted via multiple channels, require lower minimum loan amounts, are faster in processing loan applications and charge lower ATM fees. Banks in countries with more power outages require higher minimum balances for savings accounts, charge higher checking account fees, require more documents to open accounts, impose higher minimum loan amounts, take longer to process loan applications and charge higher fees for international wire transfers. 22 Theory suggests lower bank barriers in countries with more effective contractual and informational frameworks. Banks arise to overcome asymmetric information between lenders and borrowers (Diamond 1984, 1991, Ramakrishnan and Thakor 1984, Boyd and Prescott 1986), which leads to adverse selection and moral hazard problems, but the efficiency with which they are able to overcome these asymmetries, depends on the contractual and informational framework within which they operate. Specifically, more efficient systems of credit information sharing allows banks to better assess loan applicants, thus potentially reducing reliance on non- interest screening mechanisms such as minimum loan amounts and fees, while increasing the possibility to use less personal application channels such as phone or Internet and allowing for faster processing of loans. More efficient systems of contract enforcement help banks overcome problems of moral hazard and again allow them to rely less on non-interest barriers and to process loans faster. However, a more efficient contractual and information environment might also allow banks more easily to accept new deposit clients. An extensive empirical literature has shown the importance of effective contractual and informational frameworks for financial sector depth (Beck and Levine, 2005). There is empirical evidence that this relationship also holds for financial sector penetration and access to finance (Beck, Demirguc-Kunt and Martinez Peria, 2007; Beck, Demirguc-Kunt and Levine, 2005; Haselmann, Pistor and Vig, 2005; Visaria, 2006). Here we explore whether the contractual and informational frameworks have a similar importance for bank barriers. We utilize two indicators from the Doing Business database (World Bank, 2006a) that measure the efficiency of credit information systems and the cost of contract enforcement relative to GDP per capita. Our results suggest that banks in countries with more efficient systems of credit information sharing are more likely to accept deposits at multiple locations, require lower minimum balances and fewer documents to open accounts, allow for loan applications to be 23 submitted through non-traditional channels, impose lower minimum balances on consumer loans, and take less time to process SME loan applications. On the other hand, surprisingly banks in countries with better informational environments seem to charge higher fees on consumer loans and on international transfers. Banks in countries with poor systems of contract enforcement require higher minimum balances on savings accounts, charge higher fees on deposit accounts, require more documents to open accounts and impose higher minimum loan balances. The significant relationship between the efficiency of contractual and informational frameworks and lower barriers to banking thus matches the positive relationship between these institutions and aggregate financial development, established by the literature (Beck and Levine, 2005). We note, however, that it is mostly the barrier to deposit services that are significantly correlated with the contractual and informational framework rather than barriers to lending services, as one would have expected from the theoretical literature. Theory does not suggest an unambiguous relationship between market structure and barriers to banking. Banks in more concentrated banking systems might either exploit their market power imposing higher barriers or alternatively, might face higher incentives to lend to smaller, more opaque borrowers such as SMEs as they can recover investment in the relationship in future periods (Petersen and Rajan, 1995). Further, the variation of barriers across countries might be affected by the dominance of government-owned or foreign-owned banks in a banking system; banks might impose higher or lower barriers in banking systems dominated by government-owned or foreign-owned banks, independent of what individual banks' own ownership structure is. Specifically, competitive pressures or the lack thereof from a predominantly government-owned or foreign-owned banking system can push individual banks towards higher or lower banking barriers. We use data from Barth, Caprio and Levine (2004) to assess the relationship between ownership and market structure and barriers to banking. 24 Though we found that foreign banks themselves seem to charge higher fees than other banks, in foreign dominated banking systems fees are lower (perhaps because of greater competitive pressures) and it is easier to open bank accounts. On the other hand, in systems that are predominantly government-owned, customers face lower fees but also face greater restrictions in terms of where to apply for loans and the time it takes to have applications processed is longer. Finally, banks in countries with more concentrated banking systems are less likely to allow customers to open deposit accounts outside headquarters but impose lower minimum amounts for SME loans, are faster at processing loan applications and charge lower ATM fees. As recent empirical work has shown that competition is not a linear and unambiguous function of banking sector structure (Claessens and Laeven, 2004), we also explore the relationship between banking barriers and two indicators of an economy's competitiveness. First, we use the index on Banking Restrictions from the Heritage Foundation, a composite index of whether foreign banks are able to operate freely, how difficult it is to open domestic banks, what degree of regulations there are on financial market activities, the presence of state-owned banks, whether the government influences allocation of credit, and whether banks are free to provide customers with insurance products and invest in securities. Second, we use the cost of starting a formal enterprise, as share of income per capita, as an indicator of the ease of entry into the economy (World Bank, 2006a).16 Less competitive economies have banks that impose higher barriers to banking. We find that banks in economies with more restrictions to banking freedom are less likely to allow that accounts are opened outside the headquarters, demand higher minimum balances to open a checking or savings account, impose higher fees on checking accounts, require more 16We also tried regulatory indicators of bank entry, but these refer to regulatory requirements rather than the cost of setting up banks. 25 documentation to open these accounts, are less likely to accept loan applications through non- traditional channels, impose higher minimum balances on consumer loans, and are slower at processing loan applications. Banks in economies where entry into the corporate sector is more costly are less likely to allow customers to open accounts outside headquarters, charge higher fees on checking accounts, require a greater number of documents to open accounts, charge higher fees for SME loans and for using ATM cards, and take longer for processing consumer loans. More transparent societies might allow for lower barriers to banking, as banks in economies where clients have more access to information might have less leeway to impose high barriers to banking. More transparency might also imply a higher degree of competition, as customers can more easily compare products across banks. To gauge the relationship between transparency and bank barriers, we use a bank disclosure index (World Bank, 2006b) that captures how informative banks' balance sheet and income statements are. While this indicator was constructed to assess to what extent banks include information relating to their risk-taking and thus stability, it might indicate the general transparency of banking. We also utilize an indicator of media freedom, which measures the share of press outlets that are owned by the government. This indicator comes from Djankov et al. (2003), who show a negative association between this and other measures of media freedom with economic and political freedom. Our results suggest that banks in countries with higher disclosure standards require lower minimum balances to open a checking account and charge lower fees on these accounts, require fewer documents to open such an account, are more likely to accept loan applications through non-traditional channels such as phone or Internet, need fewer days to process loan applications, and charge lower fees for using ATMs. Banks in countries with lower degrees of media freedom (i.e., where a greater share is controlled by the government) restrict the locations where accounts 26 might be opened, impose higher minimum balances to open accounts, require more documents to open checking or savings accounts, need more days to process loan applications, and are less likely to accept loan applications through non-traditional channels. Overall, we find many country characteristics associated with barriers to banking services. Improvements in physical infrastructure and more efficient credit information and contract enforcement frameworks are associated with lower barriers. Barriers are also lower in countries with greater banking freedoms, greater competition and transparency. Although foreign banks themselves charge higher fees, banking systems with greater foreign entry have lower barriers in general. While government banks themselves do not seem to provide improved access, in banking systems dominated by state banks customers face lower fees but poorer quality of service (fewer locations that accept loan applications, longer loan processing times). These results have important policy implications for potential reforms to broaden access. 7. Conclusions This paper is the first effort to systematically document the existence of barriers to banking services. Using surveys of 193 banks in 58 countries, our data show significant variation in barriers to banking across countries. Though not without limitations, we think that this effort is important in identifying and understanding the channels through which financial exclusion works. Barriers like high minimum deposit balances, minimum loan amounts and fees can lead to exclusion by making these products unaffordable for large shares of the population. For example, in our sample high fees on checking and savings accounts effectively exclude more than 30 percent of the population from having a checking account in ten of our 58 countries. Also, strict documentation requirements and long processing times can exclude households and firms that cannot provide these documents or that depend on faster loan decisions. Similarly, 27 geographic centralization of deposit and loan decisions at headquarters reduces physical access and increases the opportunity costs for households and firms to access financial services. Finally, we conducted a first-cut examination of the bank and country-level factors that explain variation in indicators of bank barriers. We provide suggestive evidence that variation in these barriers is associated with variation in bank size, physical infrastructure, contractual and informational frameworks, ownership structure in banking and the general degree of competitiveness, openness and transparency of economies. While much more research is needed in this area, these results have important implications for policy reforms to broaden access. As a first attempt at capturing quantitative measures of cross-country differences in barriers to banking along the dimensions of physical access, affordability and eligibility, this paper is complementary to other efforts to collect data on access to financial services at the aggregate, firm- and household levels. We are still very much in the beginning of this work and richer data sources and in-depth analysis are needed to improve our understanding of access and its impact on economic outcomes. Going forward, several fruitful approaches can be envisioned. First, the type of analysis conducted in this paper is very useful in identifying outlier countries, i.e. those with high access barriers, as potential case studies to investigate financial access in greater depth. Case studies for individual countries that combine detailed supply and demand data from financial institutions, households and firms would be able to more thoroughly assess access to and use of financial services, barriers faced by different users, and potential policies to reduce these barriers. Compared to cross-country studies, such country-case studies can better take into account idiosyncratic characteristics and better exploit the richness of institutional detail at and below the country level. Second, while household and firm surveys at the country level are useful instruments, important empirical challenges remain in measuring the causal impact of improved access to 28 credit and deposit services on economic outcomes. Individuals and firms seeking to borrow or open bank accounts are typically different than non-borrowers, which makes causality inference from cross-sectional data very difficult. However, these identification issues may be circumvented by introducing a random component to the assignment of financial products such as subsidizing account opening fees or random variation in certain terms of the loan contract. Such randomized field experiments are likely to shed light on the impact of removing barriers on real outcomes. Third, careful cross-country studies focusing on specific standardized banking products, to the extent they exist, such as transaction accounts or consumer and SME loans would also be valuable since they allow for greater uniformity in the analysis across countries. We leave these complementary efforts for future research. 29 References Ayyagari, M., Demirguc-Kunt, A. and Maksimovic, V., (2005). How Important are Financing Constraints? The Role of Finance in the Business Environment, World Bank mimeo. Banerjee, A. and Newman, A., (1993). Occupational Choice and the Process of Development. Journal of Political Economy 101, 274-98. Barth, J., Caprio, G., and Levine, R., (2004). Bank Regulation and Supervision: What Works Best. Journal of Financial Intermediation 13, 2004, 205-248. Beck, T., Levine, R., and Loayza, N., (2000). "Finance and the Sources of Growth." Journal of Financial Economics 58, 261-300. Beck, T., Demirguc-Kunt, A. and Levine, R., (2004). Finance, Inequality and Poverty: Cross- Country Evidence. World Bank Policy Research Working Paper 3338. Beck, T., Demirguc-Kunt, A., and Levine, R., (2005). Law and Firms' Access to Finance, American Law and Economics Review 7, 211-252. Beck, T. and Levine, R., (2005). Legal Institutions and Financial Development, in: Claude Menard and Mary Shirley, eds., Handbook of New Institutional Economics, Kluwer Dordrecht (The Netherlands). Beck, T., Demirguc-Kunt, A. and Martinez Peria, M.S., (2007). Reaching Out: Access to and Use of Banking Services Across Countries, Journal of Financial Economics, forthcoming. Beck, T., Demirguc-Kunt, A. and Maksimovic, V., (2005). Financial and Legal Constraints to Firm Growth: Does Size Matter? Journal of Finance 60, 137-77. Beck, T. and de la Torre, A., (2007). The Basic Analytics of Access to Financial Services. Financial Markets, Institutions and Instruments, forthcoming. Berger, A., Hasan, I., and Klapper, L., (2004). Further Evidence on the Link between Finance and Growth: An International Analysis of Community Banks and Economic Performance, Journal of Financial Services Research 25, 169-202. Berger, A. and Udell, G., (2006). A More Complete Conceptual Framework for SME Finance. Journal of Banking and Finance 30, 2945-2966. Boyd, J. and Prescott, E., (1986). Financial Intermediary-coalitions. Journal of Economic Theory 38, 211-32 Claessens, S., (2006). Access to Financial Services: A Review of the Issues and Public Policy Issues. World Bank Research Observer, forthcoming. Claessens, S. and Demirguc-Kunt, A., (2006). Measuring Access to Financial Services through Household Level Surveys. World Bank mimeo. 30 Claessens, S., and Laeven, L. (2004). What Drives Bank Competition? Some International Evidence. Journal of Money, Credit, and Banking 36, 563-82. Demirgüç-Kunt, A. and Maksimovic, V., (1998). Law, Finance, and Firm Growth. Journal of Finance 53, 2107-2137. Diamond, D., (1984). Financial intermediation and delegated monitoring. Review of Economics Studies 51, 393-414. Diamond, D., (1991). Monitoring and reputation: the choice between bank loans and directly placed debt. Journal of Political Economy 99, 689-21. Djankov, S., McLiesh, C., Nenova, T. and Shleifer, A., (2003). Who Owns the Media? Journal of Law and Economics 46, 341-82. Dollar, D. and Kraay, A., (2002). Growth is Good for the Poor. Journal of Economic Growth 7, 195-225. Estache, A. and Goicoechea, A., (2005). A Research' Database on Infrastructure Economic Performance. World Bank Policy Research Working Paper 3643. Galor, O. and Zeira., J., (1993). Income Distribution and Macroeconomics. Review of Economic Studies 60, 35-52. Genesis (2005a). An Inter-Country Survey of the Relative Costs of Bank Accounts. Johannesburg, South Africa. Genesis (2005b). Measuring Access to Transaction Banking Services in the Southern Customs Union - an Index Approach. Johannesburg, South Africa. Gibrat, R (1931): les inégalités économiques, Paris, Sirey. Haselmann, R., Pistor, K. and Vig, V., (2005). How Law Affects Lending. Columbia Law and Economics Working Paper 285. King, R. and Levine. R., (1993). Finance, Entrepreneurship and Growth: Theory and Evidence, Journal of Monetary Economics 32, 513-42. Klapper, L., Laeven, L. and Raghuram, R., (2006). Entry Regulation as Barrier to Entrepreneurship, Journal of Financial Economics, forthcoming. Laeven, L. and Levine, R., (2006). Is there a Diversification Discount in Financial Conglomerates? Journal of Financial Economics, forthcoming. La Porta, R., Lopez-de-Silanes, F. and Shleifer, A. (2002). Government Ownership of Commercial Banks. Journal of Finance 57, 265-301. 31 Lopez, H. and Serven, L., (2006). A Normal Relationship? Poverty, Growth and Inequality, World Bank Policy Research Working Paper 3814. Mian, A. (2006). Distance Constraints: The Limits of Foreign Lending in Poor Economies. Journal of Finance 61, 1465-1505. Micco, A., Panizza, U. and Yanez, M. (2007). Bank Ownership and Performance: Does Politics Matter? Journal of Banking and Finance, forthcoming. Petersen, M. A. and Rajan, R., (1995). The Effect of Credit Market Competition on Lending Relationships. Quarterly Journal of Economics 110, 407-443. Rajan, R. and Zingales, L., (2003). Saving Capitalism from the Capitalists. Crown Business Division of Random House. Ramakrishnan, S. and Thakor, A. V., (1984). Information reliability and a theory of financial intermediation. Review of Economic Studies 51, 415-32. UNU-WIDER (2005). World Income Inequality Database, Version 2.0a. Visaria, S., (2006). The Microeconomic Impact of Debt Recovery Tribunals in India, Boston University mimeo. World Bank (2006a). Doing Business. World Bank (2006b). Global Assessment of Bank Disclosure Practices. 32 Table 1. Countries for which we collected barrier indicators Country Private GDP per Deposit Loan market Number of Credit to capita in market share banks that GDP 2000 share (respondents have USD (respondents share out of responded share out of total total system) system) 2004 2004 2004 2004 Albania 8.65% 1,477 91.42% 64.24% 5 Algeria 10.27% 1,982 34.43% 37.08% 1 Armenia 6.26% 952 59.63% 47.28% 4 Australia 100.94% 22,083 32.59% 33.59% 2 Bangladesh 27.41% 402 56.98% 56.51% 5 Belarus n.a. 1,695 74.58% 71.63% 3 Belgium 72.78% 23,213 72.56% 68.57% 3 Bolivia 42.31% 1,034 58.04% 58.87% 4 Bosnia Herzegovina n.a. 1,406 64.04% 58.96% 4 Brazil 33.89% 3,564 64.35% 48.61% 4 Bulgaria 30.86% 1,957 34.87% 31.65% 3 Cameroon 8.41% 662 83.83% 81.36% 5 Chile 70.99% 5,462 35.50% 36.05% 2 Colombia 21.80% 2,091 50.48% 45.65% 5 Croatia 54.18% 4,934 63.42% 63.69% 4 Czech Republic 30.66% 6,123 43.00% 43.00% 2 Denmark 154.04% 30,735 72.71% 48.81% 2 Dominican Republic 30.89% 2,476 39.27% 42.61% 2 Egypt, Arab Rep. 54.84% 1,615 32.05% 32.08% 2 Ethiopia 23.00% 113 93.73% 85.37% 4 France 88.19% 23,432 26.23% 30.08% 2 Georgia 8.64% 883 85.71% 80.26% 5 Germany 113.07% 23,705 31.91% 23.72% 3 Ghana 11.98% 278 69.49% 68.72% 4 Greece 72.52% 11,960 56.92% 58.36% 3 Hungary 43.65% 5,413 53.09% 42.43% 3 India 32.78% 538 36.87% 37.75% 4 Indonesia 20.99% 906 44.73% 40.38% 4 Israel 90.04% 17,788 36.17% 34.75% 2 Jordan 68.83% 1,940 83.61% 80.36% 3 Kenya 25.33% 427 43.82% 47.61% 3 n.a. means not available. 33 Table 1. Countries for which we collected barrier indicators Country Private GDP per Deposit Loan market Number of Credit to capita in market share banks that GDP 2000 USD share (respondents have (respondents share out of responded share out of total total system) system) 2004 2004 2004 2004 Korea, Rep. 125.43% 12,752 68.95% 73.54% 6 Lebanon n.a. 5,606 38.00% 38.00% 3 Lithuania 22.21% 4,402 88.87% 86.77% 5 Madagascar 8.65% 229 72.44% 74.59% 5 Malawi 8.33% 153 82.36% 59.73% 3 Malta 106.72% 9,435 44.56% 58.34% 4 Mexico 15.96% 5,968 48.95% 45.74% 3 Moldova 19.41% 400 40.16% 48.32% 3 Mozambique 2.07% 275 48.78% 40.34% 2 Nigeria 15.47% 402 32.22% 29.31% 3 Pakistan 25.74% 566 47.50% 44.02% 3 Peru 18.85% 2,206 81.88% 76.40% 4 Philippines 33.48% 1,085 41.84% 43.17% 4 Romania 8.78% 2,163 35.01% 24.66% 4 Sierra Leone 3.92% 156 100.00% 100.00% 4 Slovak Republic 30.40% 4,495 58.12% 51.93% 3 Slovenia 42.62% 10,860 67.48% 70.68% 5 South Africa 134.13% 3,312 70.09% 69.39% 3 Spain 115.46% 15,343 63.75% 66.73% 4 Sri Lanka 28.48% 962 52.19% 51.10% 3 Swaziland 14.08% 1,357 43.40% 29.19% 1 Sweden 102.82% 28,858 39.47% 22.43% 2 Switzerland 157.25% 34,340 79.57% 59.19% 2 Trinidad and Tobago 30.26% 8,055 40.15% 50.27% 3 Turkey 17.09% 3,197 50.14% 38.33% 3 Uganda 5.92% 267 59.27% 46.87% 3 Zimbabwe 16.58% 457 28.24% 43.45% 4 n.a. means not available. 34 Table 2: Indicators of barriers to accessing and using deposit services Country DEPOSITS Physical Affordability Eligibility access Locations to Minimum Minimum Minimum Minimum Annual Annual No. of docs. No. of docs open deposit amount to amount amount to amount to fees fees to open to open account open to open be be checking savings checking savings (out of 3) checking savings maintained maintained account account account account account account in checking in savings (% of (% of (out of 5) (out of 5) (% of (% of account account GDPPC) GDPPC) GDPPC) GDPPC) (% of (% of GDPPC) GDPPC) Albania 2.73 0.85 6.08 0.85 6.08 0.19 0.39 1.00 1.00 Algeria 2.00 0.00 n.a. 0.00 n.a. 0.12 0.00 3.00 1.00 Armenia 1.81 10.97 15.25 10.56 15.25 0.35 0.00 2.85 2.19 Australia 2.59 0.00 0.00 0.00 0.00 0.16 0.10 3.00 3.00 Bangladesh 2.00 2.28 0.89 2.28 0.79 0.00 0.00 4.57 4.57 Belarus 2.71 0.00 0.04 0.00 0.00 0.00 0.00 1.44 1.00 Belgium 2.00 0.00 0.00 0.00 0.00 0.09 0.00 1.80 1.80 Bolivia 2.00 17.40 0.81 25.44 3.93 0.83 1.78 2.53 2.33 Bosnia Herzegovina 2.60 0.04 0.04 0.19 0.15 0.34 0.35 1.74 1.34 Brazil 2.44 0.00 0.10 0.00 0.00 0.81 0.03 2.67 2.16 Bulgaria 2.02 0.59 0.88 0.59 0.91 0.14 0.00 1.72 1.72 Cameroon 1.88 116.39 68.26 55.88 64.75 7.87 1.22 4.00 3.11 Chile 2.42 4.33 0.00 0.00 0.00 3.38 0.42 4.42 1.58 Colombia 1.93 8.78 1.22 0.00 0.18 0.78 0.56 3.08 2.25 Croatia 2.63 0.00 1.19 0.00 0.00 0.07 0.00 2.16 2.00 Czech Republic 2.00 0.23 1.41 0.00 1.24 0.26 0.00 1.00 1.00 Denmark 2.32 0.00 0.00 0.00 0.00 0.09 0.00 1.32 1.32 Dominican Rep. 2.67 2.94 0.70 0.58 0.41 0.66 0.00 2.66 1.99 Egypt, Arab Rep. 2.00 0.35 0.00 0.18 0.18 0.40 0.07 n.a. n.a. Ethiopia 1.92 55.41 5.50 n.a. 5.11 0.00 0.00 3.77 2.14 France n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Georgia 2.56 0.00 33.18 0.00 8.09 0.33 0.33 1.66 1.78 Germany 2.65 0.00 0.01 0.00 0.00 0.26 0.00 n.a. n.a. Ghana 2.15 22.69 21.89 0.09 11.99 5.90 0.58 3.62 3.24 Greece 1.21 0.64 1.27 0.64 1.27 0.02 0.02 2.53 2.26 Hungary 2.53 0.14 2.04 0.00 0.82 0.17 0.00 1.55 1.00 India 2.00 8.85 5.02 5.83 5.02 0.00 0.17 2.69 2.55 Indonesia 2.53 9.54 3.03 6.14 0.65 2.80 0.66 3.18 2.66 Israel 2.00 0.00 0.00 0.00 0.00 0.04 0.00 1.22 n.a. Jordan 1.93 16.55 5.34 1.73 0.87 0.00 0.00 2.04 2.04 Kenya 2.78 11.71 44.30 0.00 41.82 12.82 2.07 3.78 2.86 n.a. means not available because the banks that responded to the survey account for less than 30 percent of the market 35 Table 2: Indicators of barriers to accessing and using deposit services (cont.) Country DEPOSITS Physical Affordability Eligibility access Locations to Minimum Minimum Minimum Minimum Annual Annual No. of No. of docs to open deposit amount to amount amount to amount to fees fees docs. to open savings account open to open be be checking savings open account (out of 3) checking savings maintained maintained account account checking (out of 5) account account in checking in savings (% of (% of account (% of (% of account account GDPPC) GDPPC) (out of 5) GDPPC) GDPPC) (% of (% of GDPPC) GDPPC) Korea, Rep. 2.11 3.32 0.01 0.00 0.01 0.06 0.00 1.94 1.20 Lebanon 1.58 4.22 23.98 4.22 23.98 1.96 1.90 2.54 2.36 Lithuania 2.71 0.00 1.45 0.00 1.55 0.01 0.00 1.59 1.00 Madagascar 1.95 38.86 19.35 0.00 17.59 5.15 0.00 2.94 2.71 Malawi 2.00 0.00 17.89 0.00 17.89 21.98 3.63 3.65 2.84 Malta 2.00 0.22 0.71 0.00 0.68 0.00 0.00 3.17 3.07 Mexico 2.18 1.11 0.62 0.90 0.67 0.43 0.18 2.80 2.18 Moldova 3.00 0.00 13.13 0.00 8.26 0.53 0.00 2.31 2.06 Mozambique 2.00 29.61 15.71 14.19 7.20 n.a. 0.30 1.00 1.00 Nigeria 2.44 106.42 22.07 0.00 1.96 0.05 0.00 3.66 1.99 Pakistan 2.00 1.59 1.59 0.33 0.71 0.00 0.00 2.64 2.43 Peru 2.00 1.66 0.53 0.00 0.00 1.44 0.50 2.42 1.87 Philippines 2.00 14.54 11.88 14.54 11.88 0.00 0.00 3.17 2.20 Romania 2.30 0.03 0.71 0.02 0.18 0.40 0.23 1.28 n.a. Sierra Leone 1.42 51.63 44.89 8.81 43.56 26.63 0.00 4.02 3.88 Slovak Republic 2.08 0.12 0.79 0.10 0.79 0.18 0.01 1.47 1.51 Slovenia 1.50 0.01 0.03 0.01 0.02 0.17 0.00 1.88 1.88 South Africa 2.27 0.00 1.06 0.00 0.28 2.13 0.91 3.45 3.07 Spain 1.53 0.00 0.00 0.00 0.00 0.19 0.04 1.00 1.00 Sri Lanka 1.80 15.76 3.54 4.77 0.84 0.73 0.00 2.62 1.00 Swaziland 2.00 0.00 0.48 0.00 0.48 7.24 1.09 3.00 3.00 Sweden 1.66 0.00 0.01 0.00 0.00 0.00 0.00 1.00 1.00 Switzerland 2.00 0.00 0.00 0.00 0.00 0.08 0.00 1.14 1.14 Trinidad Tobago 2.00 1.37 0.42 1.28 0.49 0.35 0.00 4.29 3.07 Turkey 2.20 0.00 0.00 0.00 0.00 0.30 0.14 3.20 2.40 Uganda 2.00 51.12 48.62 1.73 29.52 24.88 3.37 4.00 3.00 Zimbawe n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Minimum 1.21 0.00 0.00 0.00 0.00 0.00 0.00 1.00 1.00 5th percentile 1.53 0.00 0.00 0.00 0.00 0.00 0.00 1.00 1.00 Median 2.00 0.62 1.06 0.00 0.71 0.26 0.00 2.63 2.10 Average 2.14 10.93 8.14 2.94 6.15 2.43 0.38 2.54 2.09 Maximum 3.00 116.39 68.26 55.88 64.75 26.63 3.63 4.57 4.57 95th percentile 2.71 52.57 44.48 14.30 33.21 15.56 1.94 4.12 3.17 n.a. means not available because the banks that responded to the survey account for less than 30 percent of the market 36 Table 3: Indicators of barriers to accessing and using loan services Country LOANS Physical Affordability Eligibility access Locations to Minimum Fees Minimum Fees Days to Days to submit loan amount consumer amount SME process process SME applications consumer loans SME loans consumer loan (out of 5) loans (% of loans (% of loan applications (% of min. loan (% of min. loan applications GDPPC) amount) GDPPC) amount) Albania 2.03 214.29 3.45 1358.23 1.00 9.64 14.50 Algeria 2.00 45.46 0.00 0.00 0.00 8.00 30.00 Armenia 2.00 14.74 9.41 860.58 0.19 4.83 7.62 Australia 5.00 7.31 1.41 10.06 16.66 1.00 7.19 Bangladesh 2.12 25.70 0.51 9696.58 0.15 9.44 43.26 Belarus n.a. 3.28 n.a. 0.00 n.a. 8.06 6.20 Belgium 2.45 5.34 0.00 28.29 8.95 2.70 3.60 Bolivia 2.74 109.00 3.14 795.48 0.81 5.36 9.70 Bosnia Herzegovina 2.73 18.54 1.47 711.11 1.20 5.36 8.86 Brazil 4.85 1.96 5.87 8.08 2.94 1.00 3.63 Bulgaria 3.42 14.24 1.45 95.79 2.05 4.88 13.38 Cameroon 2.14 78.53 9.71 947.92 4.26 4.87 9.31 Chile 5.00 8.29 24.50 121.70 n.a. 3.84 13.87 Colombia 3.47 16.40 4.51 242.96 0.23 2.51 8.22 Croatia 3.43 3.90 1.76 22.58 0.94 2.42 4.65 Czech Republic 3.13 10.22 0.70 4.96 0.70 1.00 10.84 Denmark 5.00 0.00 2.00 0.00 1.73 0.73 1.00 Dominican Rep. 4.67 13.02 21.05 43.52 29.32 1.84 13.04 Egypt, Arab Rep. 2.81 5.84 1.65 0.00 0.88 5.38 14.43 Ethiopia 2.00 178.16 0.00 878.77 0.64 5.41 14.55 France 4.00 n.a. n.a. n.a. n.a. 4.87 10.00 Georgia 2.46 34.53 1.40 2480.08 0.99 3.31 5.62 Germany n.a. n.a. n.a. n.a. n.a. n.a. n.a. Ghana 2.63 111.94 2.86 1448.07 1.31 9.50 29.20 Greece 5.00 11.99 3.65 33.96 7.08 1.00 2.23 Hungary 3.29 4.77 3.74 58.00 3.31 5.66 7.66 India 2.44 28.79 1.19 145.17 0.93 4.17 10.75 Indonesia 3.10 31.68 n.a. n.a. n.a. 4.94 9.68 Israel 4.58 n.a. n.a. n.a. n.a. 1.00 1.79 Jordan 2.05 147.67 1.33 445.26 1.02 2.68 7.91 Kenya 3.27 186.42 1.84 166.44 1.57 2.52 5.66 n.a. means not available because the banks that responded to the survey account for less than 30 percent of the market 37 Table 3: Indicators of barriers to accessing and using loan services (cont.) Country LOANS Physical Affordability Eligibility access Locations to Minimum Fees Minimum Fees Days to Days to submit loan amount consumer amount SME process process SME applications consumer loans SME loans consumer loan (out of 5) loans (% of loans (% of loan applications (% of min. loan (% of min. loan applications GDPPC) amount) GDPPC) amount) Korea, Rep. 3.78 4.19 3.05 16.99 1.07 1.88 2.73 Lebanon 4.60 32.95 1.45 1154.76 1.29 1.58 15.61 Lithuania 4.25 6.31 2.77 17.54 0.88 2.41 8.62 Madagascar 2.16 24.06 1.43 17.27 2.46 8.55 15.46 Malawi 2.12 222.36 1.00 n.a. 1.32 1.72 n.a. Malta 4.20 19.26 3.52 355.91 0.28 1.34 5.69 Mexico 4.20 7.54 1.81 87.80 1.27 5.01 9.86 Moldova 2.54 31.11 3.34 71.78 1.34 1.36 4.31 Mozambique 2.15 30.71 n.a. 28.61 n.a. 8.66 25.84 Nigeria n.a. n.a. n.a. n.a. n.a. n.a. n.a. Pakistan 3.09 146.71 n.a. 234.25 n.a. 20.71 33.63 Peru 3.21 21.08 19.21 54.35 0.16 1.94 3.71 Philippines 2.36 330.55 1.39 916.66 n.a. 10.13 33.29 Romania n.a. n.a. n.a. n.a. n.a. n.a. n.a. Sierra Leone 1.77 143.55 2.07 243.89 1.76 1.73 9.52 Slovak Republic 3.64 10.26 n.a. 57.89 1.23 1.75 3.54 Slovenia 2.13 1.13 1.22 5.21 0.38 1.13 3.89 South Africa 5.00 7.27 4.38 15.98 1.56 1.46 4.13 Spain 5.00 9.95 1.85 19.35 1.06 1.00 1.83 Sri Lanka 2.90 36.10 0.24 20.56 n.a. 7.34 10.04 Swaziland n.a. n.a. n.a. n.a. n.a. n.a. n.a. Sweden n.a. n.a. n.a. n.a. n.a. n.a. n.a. Switzerland 3.12 0.11 0.00 11.28 0.00 1.44 3.24 Trinidad Tobago 4.62 7.71 1.33 8.30 1.24 1.33 7.32 Turkey 4.15 11.83 4.74 18.57 1.94 2.94 4.61 Uganda 2.00 205.75 2.68 3141.17 1.51 1.38 4.47 Zimbawe 2.85 24.08 3.05 240.12 2.54 1.46 3.91 Minimum 1.77 0.00 0.00 0.00 0.00 0.73 1.00 5th percentile 2.00 1.55 0.00 0.00 0.15 1.00 2.05 Median 3.11 18.54 1.84 58 1.24 2.68 8.06 Average 3.26 52.29 3.68 558 2.55 4.08 10.45 Maximum 5.00 330.55 24.50 9696.58 29.32 20.71 43.26 95th percentile 5.00 210.02 16.83 2067.28 8.67 9.56 31.48 n.a. means not available because the banks that responded to the survey account for less than 30 percent of the market 38 Table 4: Indicators of barriers to payment services PAYMENT SERVICES Affordability Country Cost to transfer Amount of fee Country Cost to transfer Amount of fee funds for using funds for using internationally ATM Cards internationally ATM Cards (% of 250 (% of 100 (% of 250 (% of 100 dollars) dollars) dollars) dollars) Albania 7.70 0.00 Korea, Rep. 7.05 0.22 Algeria n.a. 0.21 Lebanon 9.76 0.00 Armenia 6.14 0.07 Lithuania 8.72 Australia 8.05 0.00 Madagascar 4.30 0.00 Bangladesh 1.93 n.a. Malawi 6.42 0.08 Belarus 1.27 0.00 Malta 5.59 0.03 Belgium 0.12 0.00 Mexico n.a 0.40 Bolivia 13.47 0.26 Moldova 11.19 0.00 Bosnia Herzegovina 3.79 0.01 Mozambique n.a n.a Brazil 14.85 0.11 Nigeria n.a 0.50 Bulgaria 5.24 0.13 Pakistan n.a 0.60 Cameroon 9.15 0.00 Peru 6.68 0.24 Chile n.a. 0.00 Philippines n.a 0.00 Colombia n.a. 0.19 Romania n.a. n.a. Croatia 3.57 0.00 Sierra Leone 6.86 0.00 Czech Republic 3.99 0.19 Slovak Rep. 4.38 0.19 Denmark 4.09 0.00 Slovenia 2.88 0.00 Dominican Republic 20.00 n.a. South Africa 9.53 0.34 Egypt, Arab Rep. 0.76 0.00 Spain 6.39 0.00 Ethiopia 1.87 0.00 Sri Lanka n.a. n.a. France n.a. n.a. Swaziland 14.40 n.a. Georgia 7.03 0.13 Sweden 8.16 0.00 Germany n.a. n.a. Switzerland 3.17 0.00 Ghana 14.70 0.19 Trinidad and Tobago 3.74 0.05 Greece 7.42 0.00 Turkey 6.34 0.00 Hungary 3.60 n.a. Uganda 0.55 0.19 India 6.49 0.00 Zimbawe n.a. n.a. Indonesia 2.83 0.00 Israel n.a. 0.23 Minimum 0.12 0.00 Jordan 5.37 0.00 5th percentile 0.83 0.00 Kenya 8.43 0.15 Median 6.36 0.00 Average 6.54 0.10 Maximum 20.00 0.60 95th percentile 14.66 0.38 n.a. means not available because the banks that responded to the survey account for less than 30 percent of the market 39 Table 5: Correlations between indicators of barriers * significant at 10%; ** significant at 5%, *** significant at 1%. of to to to to Fees ) ) Loan 5) E 250) of Open (Out Loan) Loan) SM of Account Account Account ount 5) Account 5) to out( Am ansoL ount s Loans Am an mu Loans mu ocess ansoL Funds er %( 3) Balance Balance Savings of of Submit Fees er Lo E Pr er ocess Account mu DPPC)Gfo mu Checking GDPPCfo GDPPCfo Documents Documents of of to um um nimi SM nimi to Pr Applications ansfrT Checking Savings DPPC)Gfo utO( Savings utO( DPPC)Gfo DPPC)Gfo SME Consumer M M to Locations nimi (% nimi (% (% (% Checking (% (% Applications to Annual nimi nimi of Fees of M M M Consum M Days Consum Days ansoL ationallyn Deposit Open Open Annual Number Open Number Open Applications Fees %( %( Locations Cost ternI Minimum Balance to Open Checking Account -0.013 (% of GDPPC) Minimum Balance to Open Savings Account -0.054 0.748*** (% of GDPPC) Annual Checking Fees (% of GDPPC) -0.134* 0.242*** 0.361*** Annual Savings Fees (% of GDPPC) -0.06 0.011 0.056 0.315*** Number of Documents to Open Checking Account -0.118 0.196*** 0.163** 0.283*** 0.187** (Out of 5) Number of Documents to Open Savings Account -0.109 0.217*** 0.205*** 0.272*** 0.266*** 0.779*** (Out of 5) Locations to Submit Loan Applications 0.149** -0.148** -0.153** -0.217*** -0.046 -0.085 -0.13* (out of 5) Minimum Amount Consumer Loans (% of -0.052 0.246*** 0.391*** 0.226*** 0.085 0.157** 0.159** -0.231*** GDPPC) Minimum Amount SME Loans (% of GDPPC) 0.005 0.072 0.13* 0.03 0.023 0.063 0.09 -0.113 0.098 Fees Consumer Loan (% of Minimum Loan 0.033 -0.059 -0.043 -0.014 -0.019 0.027 -0.14* 0.028 -0.107 -0.064 Amount) Fees SME Loan (% of Minimum Loan Amount) 0.064 -0.037 -0.031 0.023 0.006 -0.002 -0.025 0.215*** -0.085 -0.061 0.265*** Days to Process Consumer Loan -0.119 0.095 0.067 -0.048 -0.108 0.124* 0.122 -0.252*** 0.404*** 0.095 -0.041 -0.077 Applications Days to Process SME Loan Applications -0.12 0.021 0.065 0.06 -0.073 0.214*** 0.223*** -0.282*** 0.301*** 0.387*** -0.087 -0.066 0.65*** Cost to Transfer Funds Internationally (% of 250) 0.069 -0.094 0.027 0.059 0.037 -0.008 -0.008 0.134* -0.016 -0.068 0.219*** 0.214*** -0.023 -0.033 Fees for Using ATM Card -0.026 0.039 -0.011 0.069 0.076 0.168** 0.182** -0.019 -0.017 -0.025 -0.025 -0.067 -0.039 -0.056 0.267*** 40 Table 6. Correlations between barriers indicators and measures of financial and economic development and financial outreach * significant at 10%; ** significant at 5%, *** significant at 1%. Indicator GDP per Private Number of Number of Number of Business Business capita Credit / branches Loans Per Deposits per constraint: constraint: GDP per 100,000 1000 People 1000 People access to cost of people finance finance Locations to Open a Deposit Account -0.106 -0.077 -0.199 -0.394* -0.245 0.011 0.033 (Out of 3) Minimum Balance to Open Checking Account -0.283** -0.319** -0.291** -0.342 -0.467*** 0.371** 0.381** (% of GDPPC) Minimum Balance to Open Savings Account -0.33** -0.446*** -0.292** -0.315 -0.432** 0.403** 0.482*** (% of GDPPC) Annual Checking Fees -0.248* -0.297** -0.213 -0.202 -0.317* 0.368** 0.513*** (% of GDPPC) Annual Savings Fees -0.255* -0.219 -0.231 -0.264 -0.405** 0.285* 0.415** (% of GDPPC) Number of Documents to Open Checking Account -0.438*** -0.289** -0.392*** -0.181 -0.411** 0.4** 0.272 (Out of 5) Number of Documents to Open Savings Account -0.306** -0.201 -0.273* -0.168 -0.346* 0.36** 0.319* (Out of 5) Locations to Submit a Loan Application 0.433*** 0.564*** 0.421*** 0.67*** 0.466*** -0.378** -0.378** (Out of 5) Minimum SME Loan Amount -0.203 -0.193 -0.187 -0.224 -0.28 0.263 0.212 (% of GDPPC) Minimum Consumer Loan Amount -0.348** -0.325** -0.29* -0.387* -0.452** 0.203 0.253 (% of GDPPC) Fees for SME Loan 0.159 0.088 0.111 -0.016 0.115 -0.091 -0.034 (% of Minimum Loan Amount) Fees for Consumer Loan -0.157 -0.052 -0.129 0.077 -0.132 0.015 -0.048 (% of Minimum Loan Amount) Days to Process Consumer Loan Application -0.355*** -0.369** -0.301** -0.413* -0.418** 0.226 0.19 Days to Process SME Loan Application -0.372*** -0.394*** -0.313** -0.412* -0.405** 0.318* 0.313* Cost to Transfer Funds Internationally -0.192 -0.103 -0.116 -0.091 -0.278 -0.002 0.127 (% of 250) Fees for using ATM cards -0.279** -0.32** -0.3** -0.263 -0.305 0.109 0.171 41 Table 7. Back-of-the envelope calculations of the share of the population that cannot afford deposit accounts Average HH Checking Savings GDP per capita Gini Coefficient Lowest percentile for which Size Account Account (in 2003 USD) (latest available year) fee is more than 2% of HH Annual Fee Annual Fee income (in 2003 USD) (in 2003 USD) Checking Savings Account Fee Account Fee Albania 4.24 3.44 7.06 1811.11 0.28 1 1 Algeria 4.85 2.56 0.00 2134.54 0.35 1 0 Armenia 4.12 3.23 0.00 924.23 0.36 1 0 Australia 3.84 42.46 26.54 26539.40 0.31 1 1 Bangladesh 4.80 0.00 0.00 380.00 0.32 0 0 Belarus 0.00 0.00 1805.30 0.25 Belgium 2.56 26.39 0.00 29320.13 0.29 1 0 Bolivia 4.18 7.60 16.30 915.90 0.53 5 17 Bosnia and Herzegovina 6.16 6.34 1811.88 0.26 Brazil 3.79 22.58 0.84 2787.90 0.61 12 1 Bulgaria 2.71 3.57 0.00 2548.76 0.37 1 0 Cameroon 5.17 68.32 10.59 868.16 0.44 54 3 Chile 3.44 156.16 19.40 4620.02 0.57 48 4 Colombia 4.78 14.01 10.06 1795.65 0.57 6 4 Croatia 3.00 4.54 0.00 6484.10 0.31 1 0 Czech Republic 2.43 23.09 0.00 8880.78 0.23 1 0 Denmark 2.18 35.26 0.00 39181.91 0.35 1 0 Dominican Republic 3.90 12.47 0.00 1889.28 0.48 2 0 Egypt, Arab Rep. 4.67 4.65 0.81 1163.56 0.38 1 1 Ethiopia 4.83 0.00 0.00 115.75 0.30 0 0 France 2.53 29805.15 0.27 Georgia 3.52 2.89 2.89 874.42 0.45 1 1 Germany 2.29 76.97 0.00 29602.50 0.28 1 0 Ghana 5.11 21.21 2.08 359.43 0.41 37 1 Greece 2.99 3.14 3.14 15700.09 0.32 1 1 Hungary 2.67 13.95 0.00 8208.52 0.27 1 0 India 5.31 0.00 0.96 564.32 0.26 0 1 Indonesia 3.97 30.97 7.30 1105.94 0.34 10 1 Israel 3.50 6.60 0.00 16493.07 0.37 1 0 Jordan 6.16 0.00 0.00 1978.74 0.36 0 0 Kenya 4.55 58.89 9.51 459.35 0.45 81 10 Korea, Rep. 4.41 7.63 0.00 12709.67 0.37 1 0 42 Table 7. Back-of-the envelope calculations of the share of the population that cannot afford deposit accounts (cont.) Average HH Checking Account Savings GDP per Gini Coefficient Lowest percentile for Size Annual Fee Account capita (latest available which fee is more than (in 2003 USD) Annual Fee (in 2003 year) 2% of HH income (in 2003 USD) USD) Checking Savings Account Account Fee Fee Lebanon 111.77 108.35 5702.64 0.60 Lithuania 2.57 0.54 0.00 5369.39 0.36 1 0 Madagascar 4.89 15.99 0.00 310.57 0.47 40 0 Malawi 4.37 31.43 5.19 143.01 0.49 94 33 Malta Mexico 4.38 27.20 11.39 6326.51 0.51 2 1 Moldova 2.48 0.00 468.16 0.44 Mozambique 4.43 0.75 251.18 0.39 1 Nigeria 4.97 0.23 0.00 462.98 0.50 1 0 Pakistan 6.80 0.00 0.00 554.77 0.31 0 0 Peru 32.23 11.19 2238.11 0.49 Philippines 5.31 0.00 0.00 1004.02 0.50 0 0 Romania 3.13 10.95 6.30 2736.97 0.29 1 1 Sierra Leone 6.76 51.48 0.00 193.32 0.64 89 0 Slovak Republic 10.93 0.61 6071.99 0.27 Slovenia 3.07 23.91 0.00 14064.90 0.22 1 0 South Africa 4.00 77.23 33.00 3625.87 0.60 31 12 Spain 3.28 39.85 8.39 20974.39 0.31 1 1 Sri Lanka 3.84 6.92 0.00 947.72 0.47 2 0 Swaziland 5.39 124.85 18.80 1724.49 0.60 61 10 Sweden 2.04 0.00 0.00 33670.48 0.26 0 0 Switzerland 2.42 35.08 0.00 43847.96 0.17 1 0 Trinidad and Tobago 3.68 29.04 0.00 8296.73 0.40 1 0 Turkey 5.05 10.20 4.76 3399.36 0.40 1 1 Uganda 4.86 57.92 7.84 232.79 0.55 93 33 Zimbabwe 4.81 615.20 0.73 43 Table 8. Bank-level Regression Results Table shows results of regressing each indicator against the four bank-level variables (two ownership dummies, loan to assets and log of assets) along with one country level variable at a time. Regressions are estimated via OLS in all cases except for regressions on the number of places to open a deposit account and the number of places to submit a loan application where ordered probit models are estimated. Robust standard errors in brackets. * significant at 10%; ** significant at 5%, *** significant at 1%. se ed ut ed ut sit en fo es er en fo er Op Op of Fet Fe (Ot (O ) oanL nt) 0) of nt) Depon 3)fo un un 5) oanL toe (%t toe %( ) unt ) edeN edeN unt E rdaC co of ) %(s ns E SM an ns sdnuF 25fo ut un ) unt ) Ac cocA ntse ccoA ntse oanLit onsumC 0) Opeot (Ot ncla bm co ncla )5 )5 out(n nt Ac of of PPCDGfo nt PPCDG Lo (%naoL onsumCs tioac SMs es es tioac sfern (%yl MTAg un Ba Ba cocA ngki PPCDG PPCDG 10fo of of ngkic cocAs Suot of ermu oumAnaoL oumAnaoL oc plipA oc Tra nsio m ccoA mui ngikce PPCDG m gs PPCDG ec cumoD cumoD tioac oumA (% oumA ns mu SME mu ppliA to sinUr %( mui vin foe Locat Min Ch Sa Min Chlaun (% gsnivaS (% of of nsio al um um (% es Prot Prot st ber heCn ber ppliA anoL nimi Fe nimi anoL Co nalioatnre An nnuA umN peO to umN ngviaSnepO ocatL nimi nimi M M aysD aysD Int Fe to M M CoseeF Bank-level Govt. Ownership 0.03 0.034 0.002 -0.139 -0.04 0.042 -0.057 -0.275 0.04 -0.584 0.003 0.164 0.543 4.973** -0.162 -0.02 Dummy [0.255] [0.214] [0.172] [0.101] [0.038] [0.065] [0.063] [0.253] [0.377] [0.529] [0.197] [0.151] [0.736] [2.427] [0.164] [0.039] Bank-level Foreign Ownership -0.12 -0.088 0.338 0.704*** 0.162* 0.053 0.06 0.002 0.447 0.018 0.357** 0.143 -0.589 -0.217 0.102 0.027 Dummy [0.225] [0.283] [0.262] [0.214] [0.095] [0.071] [0.066] [0.219] [0.284] [0.540] [0.147] [0.127] [0.758] [1.790] [0.164] [0.036] Bank-level Loans / Assets 0.087 -0.752 -1.268** -0.162 -0.117 0.323* 0.095 1.113** -0.37 0.843 -0.072 -0.271 3.195* 1.329 -0.265 -0.023 [0.589] [0.604] [0.504] [0.364] [0.165] [0.176] [0.163] [0.557] [0.780] [1.250] [0.414] [0.310] [1.701] [4.238] [0.349] [0.093] Bank-level Log(Total -0.012 -0.208*** -0.217*** -0.119*** -0.022** -0.031*** -0.025** 0.239*** -0.276*** -0.283*** -0.006 -0.007 -0.408*** -0.818*** 0.029 -0.002 Assets) [0.043] [0.040] [0.031] [0.024] [0.009] [0.010] [0.010] [0.040] [0.055] [0.081] [0.023] [0.021] [0.114] [0.229] [0.027] [0.005] Tel. lines per capita 0 -0.004*** -0.002*** -0.002*** -0.000*** -0.001*** -0.000*** 0.001* -0.003*** -0.003** 0 0 -0.007*** -0.018*** 0 -0.000** [0.001] [0.001] [0.000] [0.000] [0.000] [0.000] [0.000] [0.001] [0.001] [0.001] [0.000] [0.000] [0.002] [0.005] [0.000] [0.000] Electric Power Transmission and Distribution 0.016 0.015 0.040*** 0.014** 0 0.012*** 0.009*** -0.001 0.055*** 0.060** 0.020*** 0.005 0.097* 0.178* 0.028*** 0.003 Losses (% of [0.012] [0.016] [0.013] [0.007] [0.002] [0.003] [0.003] [0.010] [0.018] [0.026] [0.008] [0.005] [0.052] [0.107] [0.007] [0.002] output) Credit Information Index 0.116** -0.102 -0.180*** -0.024 -0.013 -0.036** -0.039** 0.141*** -0.126* -0.193 0.075* 0.01 -0.261 -0.833* 0.101*** 0.007 [0.057] [0.068] [0.056] [0.046] [0.019] [0.016] [0.015] [0.052] [0.074] [0.120] [0.040] [0.029] [0.187] [0.449] [0.037] [0.010] Costs of Enforcing Contracts -0.001 0.008 0.012** 0.012*** 0.005** 0.004*** 0.003*** -0.006 0.012* 0.018 -0.002 0 0.004 0.031 0.001 0 (% of debt) [0.004] [0.008] [0.005] [0.004] [0.002] [0.001] [0.001] [0.006] [0.007] [0.015] [0.002] [0.002] [0.012] [0.050] [0.002] [0.000] 44 Table 8. Bank-level Regression Results Table shows results of regressing each indicator against the four bank-level variables (two ownership dummies, loan to assets and log of assets) along with one country level variable at a time. Regressions are estimated via OLS in all cases except for regressions on the number of places to open a deposit account and the number of places to submit a loan application where ordered probit models are estimated. Robust standard errors in brackets. * significant at 10%; ** significant at 5%, *** significant at 1%. ut n n of (O oanL s on %( unt ngki of (% fo lly um an cocA hecCn ngsvi of (% peOot peOot nioat lic anoL Lo atiic (% eeF ed m n sit San (%e )5 ed er M oaL pplA al Fe edeN of edeN 5) App anoL nimi poeDnepOot peOot er ) ) of E of nnuA oanL naioatnretnI ce analB PCPDG peOot ce nt alunnA ntse ntse ut Loan onsumC SM (% Minimufo E nds rdaC (O bmit nt nt of analB PCPDG of unt cumoD cumoD unt Suot oumA oumA anoL er %(na onsumCs SMs es es Furef MTAg nsio um (%t um (%t ouccAg Lo oc ns oc ans ) un un kin cocAsg ) of utO(tnuoccAg of ber kin ber cocAsg um nsio 5) mu ) mu ) ) E ) Prot tioac Prot Tr to 0) sinUr 0) catoL )3 nimi 10fo of M ccoA nimi M ccoA hecC PPCDG vin Sa PPCDG umN echC vin oft umN nimi Sa Locat (ou M PPCDG nimi M PPCDG onsCe Fe ountmA MSeeF ountmA aysD ppliA ysaD stoC 25fo foe Fe %( Govt. Bank Share -0.004 -0.004 -0.001 -0.004* 0 -0.003* -0.003* -0.020*** 0.012 -0.001 0.002 -0.004 0.044* 0.063 0.005 -0.001 [0.005] [0.007] [0.007] [0.002] [0.001] [0.002] [0.002] [0.006] [0.008] [0.015] [0.004] [0.003] [0.023] [0.052] [0.004] [0.001] Foreign Bank Share 0.010* -0.008* -0.007 -0.006** -0.001 -0.004*** -0.005*** 0 -0.006 0.013 0 0.004* 0.017 -0.02 -0.003 0 [0.006] [0.004] [0.005] [0.003] [0.001] [0.001] [0.001] [0.005] [0.006] [0.008] [0.002] [0.002] [0.015] [0.033] [0.003] [0.001] Bank Concentration -1.541*** -0.547 -0.353 0.456 -0.119 -0.158 -0.041 0.013 -1.434 -3.453*** -0.452 -0.125 -4.232** -12.207*** -0.297 -0.163** [0.543] [0.710] [0.539] [0.373] [0.184] [0.159] [0.155] [0.606] [0.878] [1.300] [0.387] [0.296] [1.845] [4.322] [0.375] [0.081] Index of Banking -0.260** 0.325** 0.364*** 0.221*** 0.016 0.127*** 0.083*** -0.365*** 0.248* -0.033 0.015 0.025 0.726* 1.811** 0.028 0.034 Restrictions [0.101] [0.128] [0.099] [0.082] [0.034] [0.027] [0.028] [0.104] [0.148] [0.236] [0.073] [0.064] [0.386] [0.855] [0.072] [0.023] Costs of Starting a Business -0.001** 0.001 0.001 0.002*** 0 0.000*** 0.000*** 0 0.001 0.001 0 0.000*** -0.002** -0.004 0 0.000*** (% of income [0.000] [0.001] [0.001] [0.000] [0.000] [0.000] [0.000] [0.000] [0.001] [0.001] [0.000] [0.000] [0.001] [0.003] [0.000] [0.000] per capita) Bank Disclosure Index 0.013 -0.026** -0.01 -0.022*** -0.002 -0.009*** -0.004* 0.021** -0.003 0.016 -0.005 -0.005 -0.061** -0.138** 0.006 -0.003* (Composite) [0.010] [0.010] [0.009] [0.006] [0.002] [0.003] [0.002] [0.008] [0.013] [0.020] [0.006] [0.005] [0.029] [0.059] [0.006] [0.002] Share of Govt. Owned Media -0.871*** 1.654*** 0.581* 0.395 0.055 0.252*** 0.165* -0.783*** 0.386 0.4 0.01 -0.068 2.094** 5.809** -0.397 -0.062 [0.284] [0.420] [0.342] [0.252] [0.121] [0.084] [0.085] [0.299] [0.552] [0.853] [0.238] [0.175] [1.056] [2.620] [0.240] [0.040] 45 APPENDIX Technical appendix for section 5 The use of a lognormal function to model income distribution was first suggested by Gibrat (1931) and widely used in the subsequent literature. Recently, Lopez and Serven (2006) show that the size distribution of income per capita is indeed very well approximated by a lognormal density function. Specifically, they cannot reject the null hypothesis that theoretical income quintiles shares computed from the Gini coefficient are equal to empirically observed quintile shares from income-based household surveys. Log normality implies the following relationship between the Gini coefficient G, the standard deviation of log income and the Lorenz curve L(p): = 2 -1 [(1+G)/2] (1) L(p) = (-1(p) - ) (2) where (.) denotes the cumulative normal distribution. The assumption of log-normality thus implies a one-to-one mapping of the Gini coefficient and the Lorenz curve and therefore also a one-to-one mapping between the Gini coefficient and income percentiles. We therefore can use the observed Gini coefficient to calculate theoretical income percentiles Pj j = 1,...,99 as follows: Pj = L(.01j) ­ L(.01(j-1)) j=1,...,99. (3) Substituting in (1) and (2) yields: Pj = {-1(.01j) - 2 -1 [(1+G)/2]} - {-1(.01(j-1)) - 2 -1 [(1+G)/2]} (4) We can then compute income per capita yj for each percentile j as function of Pj and income per capita y. yj = yPj/0.01. (5) We then multiply yj with household size to get to the average household income hj for each income distribution percentile. While household size is expected to vary with income level within countries, we do not have data available on household size distribution, and are therefore not able to adjust for this effect. Finally, we compare hj j=1,...,99 with the annual checking and saving account fee to determine j such that 0.02*hj < account fee and .02*hj+1 > account fee. Income distribution percentile j thus indicates the percentage of the population that cannot afford checking (saving) account services. Data on income per capita and household size are from World Development Indicators and Gini data are from UNU-WIDER (2005). 46 Table A.1: Barriers to accessing and using business and mortgage loans Country LOANS Physical Affordability Eligibility access No. of places Minimum Fee Minimum Fee Days to Days to to submit amount business amount mortgage process process loan business loan (% mortgage loan (% business mortgage applications loan (% of min. loan (% of min. loan loan (out of 5) of loan of loan applications applications GDPPC) amount) GDPPC) amount) Albania 2.03 2263.77 1.00 535.19 2.25 16.05 11.69 Algeria 2 38967.05 0.00 1298.9 0.30 14 30 Armenia 2 1042.28 0.19 234.16 2.81 9.94 10.95 Australia 5 10.06 16.66 41.12 1.35 7.19 2.59 Bangladesh 2.12 55.28 0.15 1412.52 0.18 34.55 33.48 Belarus n.a. 7.12 n.a. 0 n.a 7.34 8.74 Belgium 2.45 28.29 8.95 86.18 1.36 3.6 5.24 Bolivia 2.74 759.35 0.81 1124.84 0.59 23.26 15.03 Bosnia Herzegovina 2.73 573.97 1.20 484.92 1.49 14.7 16.65 Brazil 4.85 19.19 2.94 n.a. n.a 10.32 n.a. Bulgaria 3.42 130.35 2.05 213.32 11.41 21.38 6.84 Cameroon 2.14 16393.68 4.26 1544.77 0.86 12.91 16.97 Chile 5 n.a. n.a. 213.2 0.34 n.a. 70.63 Colombia 3.47 2131.83 0.23 n.a. n.a 11 5.14 Croatia 3.43 146.24 0.94 183.04 1.17 11.89 4.53 Czech Republic 3.13 4.96 0.70 84.65 0.60 8.05 6.66 Denmark 5 0 1.73 0 1.59 1 4.56 Dominican Rep. 4.67 89.32 29.32 176.1 3.56 6.67 17.55 Egypt, Arab Rep. 2.81 14.61 0.88 0 0.49 19.29 38.72 Ethiopia 2 981.67 0.64 712.65 0.68 14.55 15 France 4 n.a. n.a. n.a. n.a 18.22 24.67 Georgia 2.46 2345.59 0.99 290.71 0.73 5.03 4.56 Germany n.a. n.a. n.a. n.a. n.a n.a. n.a. Ghana 2.63 1044.39 1.31 1320.35 2.01 19.07 n.a. Greece 5 13.98 2.02 80.86 10.63 4.77 5.43 Hungary 3.29 58 3.31 29 2.78 10.04 19.94 India 2.44 57.77 0.93 145.17 0.74 19.98 9.45 Indonesia 3.1 n.a. n.a. n.a. n.a 16.59 6.07 Israel 4.58 n.a. n.a. n.a. n.a 1.79 12.08 Jordan 2.05 354.7 1.02 362.27 0.85 8.16 7.24 Kenya 3.27 193.78 1.57 n.a. n.a 5.66 n.a. n.a. means not available because the banks that responded to the survey account for less than 30 percent of the market 47 Table A.1: Barriers to accessing and using business and mortgage loans (cont.) Country LOANS Physical Affordability Eligibility access No. of places Minimum Fee Minimum Fee Days to Days to to submit amount business amount mortgage process process loan business loan (% mortgage loan (% business mortgage applications loan (% of min. loan (% of min. loan loan (out of 5) of loan of loan applications applications GDPPC) amount) GDPPC) amount) Korea, Rep. 3.78 16.99 1.07 4.19 5.35 2.73 2.36 Lebanon 4.6 4470.83 1.29 409 2.04 15.61 9.26 Lithuania 4.25 17.54 0.88 65.83 0.80 9.83 8.48 Madagascar 2.16 17.27 2.46 n.a. n.a 18.6 n.a. Malawi 2.12 306.05 1.32 1738.08 1.14 15.39 14.16 Malta 4.2 529 0.28 275.38 0.27 5.64 2.74 Mexico 4.2 101.93 1.27 298.56 1.40 15.7 28.25 Moldova 2.54 64216.77 1.34 428.58 1.09 7.31 3.9 Mozambique 2.15 28.61 n.a. 71.53 n.a 25.84 34.21 Nigeria n.a. n.a. n.a. n.a. n.a n.a. n.a. Pakistan 3.09 n.a. n.a. 954.59 n.a 31.98 28.44 Peru 3.21 429.43 0.16 410.39 2.58 10.63 3.81 Philippines 2.36 920.23 n.a. 763.35 1.04 44.13 12.21 Romania n.a. n.a. n.a. n.a. n.a n.a. n.a. Sierra Leone 1.77 218.23 1.76 5157.4 1.00 11.53 4.66 Slovak Republic 3.64 50.91 1.23 71.15 n.a 3.06 4.67 Slovenia 2.13 5.21 0.38 94.9 1.30 4.19 7.6 South Africa 5 15.98 1.56 142.37 1.00 2.73 5.55 Spain 5 19.35 1.06 100.19 0.89 1.83 3.22 Sri Lanka 2.9 20.56 n.a. 51.64 1.00 15.57 20.61 Swaziland n.a. n.a. n.a. n.a. n.a n.a. n.a. Sweden n.a. n.a. n.a. n.a. n.a n.a. n.a. Switzerland 3.12 11.28 0.00 22.57 0.00 3.24 1.56 Trinidad Tobago 4.62 8.3 1.24 93.03 1.02 10.41 7.5 Turkey 4.15 74.26 1.94 n.a. 2.00 13.75 n.a. Uganda 2 7039.03 1.51 n.a. n.a 5.15 n.a. Zimbabwe 2.85 263.49 2.54 n.a. n.a 7.91 n.a. Minimum 1.77 0 0.00 0 0.00 1 1.56 5th percentile 2 5.88 0.15 0.42 0.27 2.32 2.63 Median 3.11 95.62 1.24 213.2 1.06 10.52 8.61 Average 3.26 3051.43 2.43 505.27 1.82 12.3 13.34 Maximum 5 64216.77 29.32 5157.4 11.41 44.13 70.63 95th percentile 5 13119.55 8.25 1531.55 5.61 28.61 34.02 n.a. means not available because the banks that responded to the survey account for less than 30 percent of the market 48 Table A.2: Definition and sources for explanatory variables in Table 8 Variable Source Bank-level Government Ownership Dummy Bank-level Foreign Ownership Dummy Micco, Panizza, andYanez (2007) Bank-level Loans / Assets Bank-level Total Assets BankScope Database (August 2006). Fitch Ratings/Bureau van Dijk Bank Concentration Tel. lines per capita Electric Power Transmission and Distribution Losses (% of output) Estache and Goicoechea. (2005) Credit Information Index Costs of Enforcing Contracts (% of debt) World Bank (2006a) Costs of Starting a Business (% of income per capita) Govt. Bank Share Foreign Bank Share Barth, Caprio, Levine. (2004). Index of Banking Restrictions Index of Economic Freedom 2006. The Heritage Foundation/The Wall Street Journal Bank Disclosure Index (Composite) World Bank (2006b) Share of Media Outlets Owned by the Government Djankov et al. (2003) 49 Table A.3. Summary Statistics for Explanatory Variables in Table 8 Variables Obs. Std. Dev. Mean Min Median Max Bank-level Government Ownership Dummy 177 0.40 0.19 0.00 0.00 1.00 Bank-level Foreign Ownership Dummy 177 0.41 0.21 0.00 0.00 1.00 Bank-level Loans / Assets 185 0.16 0.46 0.00 0.48 0.80 Bank-level Log of Total Assets 185 2.48 14.72 9.23 14.51 21.06 Tel. lines per capita 58 215.59 223.63 2.35 174.17 745.31 Electric Power Transmission and Distribution Losses (% of output) 53 9.59 14.83 3.06 13.16 49.89 Credit Information Index 55 2.02 3.20 0.00 4.00 6.00 Costs of Enforcing Contracts (% of debt) 55 23.74 23.42 5.20 16.20 136.50 Govt. Bank Share 47 23.97 18.60 0.00 11.00 96.00 Foreign Bank Share 43 29.63 38.42 0.00 30.00 90.00 Bank Concentration 57 0.17 0.66 0.34 0.64 1.00 Index of Banking Restrictions 58 1.02 2.66 1.00 3.00 5.00 Costs of Starting a Business (% of income per capita) 55 220.01 78.65 0.00 22.40 1442.50 Bank Disclosure Index (Composite) 58 12.49 63.72 28.00 63.00 89.00 Share of Media Outlets Owned by the Government 46 0.34 0.20 0.00 0.00 1.00 50 Table A.4. Correlation between Explanatory Variables in Table 8 * significant at 10%; ** significant at 5%, *** significant at 1%. a ndexI e n Assets/ debt) e Businessa capita) Indexe tletsuO ntemn ntemn capitr r and Losses ation of Shar atio per Gover igne pe Powec ission output) rmofnI ingcrofnE %( Shar Banking the Gover Dummy For Dummy Loans Bank Starting Media by of lines acts Bank Concentr of ctionsi of income Disclosur posite) of level level Tel. Electri ansmrT of edit igne of Distribution %( Cr Costs Contr Govt. For Bank ndexI Restr Costs %( Bank Com( Share Owned Bank- Ownership Bank- Ownership Bank-level Electric Power Transmission and Distribution Losses (% of output) -0.568*** Credit Information Index 0.469*** -0.482*** Costs of Enforcing Contracts (% of debt) -0.438*** 0.292*** -0.25*** Govt. Bank Share -0.08 0.195** -0.157* 0.253*** Foreign Bank Share -0.116 0.072 -0.109 -0.236*** -0.372*** Bank Concentration 0.054 -0.368*** -0.198*** 0.054 -0.319*** 0.179** Index of Banking Restrictions -0.478*** 0.398*** -0.463*** 0.426*** 0.567*** -0.4*** -0.079 Costs of Starting a Business (% of income per capita) -0.291*** 0.142* -0.363*** 0.076 -0.098 -0.082 0.243*** 0.443*** Bank Disclosure Index (Composite) 0.577*** -0.482*** 0.404*** -0.184** 0.062 -0.297*** -0.124* -0.382*** -0.37*** Share of Media Outlets Owned by the Government -0.422*** 0.207** -0.569*** -0.037 0.161* 0.052 0.284*** 0.312*** 0.282*** -0.463*** Bank-level Government Ownership Dummy -0.082 0.09 -0.039 0.058 0.561*** -0.272*** -0.214*** 0.364*** -0.002 0.036 -0.025 Bank-level Foreign Ownership Dummy -0.1 -0.086 -0.026 0.056 -0.244*** 0.471*** 0.23*** -0.213*** 0.066 -0.123 0.067 -0.251*** Bank-level Loans / Assets 0.101 -0.164** 0.204*** -0.2*** -0.053 -0.059 -0.164** -0.031 -0.165** 0.071 0.004 -0.131* -0.136* Bank-level Log of Total Assets 0.606*** -0.475*** 0.542*** -0.202*** 0.154* -0.467*** -0.275*** -0.134* -0.263*** 0.589*** -0.404*** 0.223*** -0.171** 0.051 51 Figure 1. Locations to Open a Deposit Accounts Moldova Republic 3 rzegovinaeH nyaeK banialA lBe rusa y and Lithuania minicanoD ian ai German oatiarC Bos Australia orgiaeG ungary Indones H eria 3) Nig azilrB ilehC arkm ani ricafA Tobago Rep. of Den Roma outhS a p.eR Republic Turkey an and Mexico rabA Republic car Gh Korea, (Out ovaklS lgariauB andagU TrinidadSwitzerland Swaziland P ilippinesh stani rael an Peru Pak Mozambique Malta Malawi Is India Egypt, Czech liviaoB lgiumeB ngladeshaB eria Alg Madagas 2 lombiaoC drJo ai Ethiopia Cameroon ArmenLankairS Account Sweden Lebanon ain Leone Sp ovenialS Sierra Deposit Greece Open 1 to Location 0 Sample size: 56 countries Figure 2. Minimum Balance to Open Checking Account (% of GDPPC) 150 GDPPC) of meroonaC ai (% ger Ni 100 Account Checking one Le Ethiopia rra Open Sie Uganda car to 50 Madagas Mozambique Balance Ghana na rd Lanka a ai publiceR Tobago Bolivia Jo Sri Philippines meni p.eR h p.eR Herzegovina and Kenya Ar and Indones ades n Arab publiceR nd k India Colombia ile ea, o iar publiceR nia a ain y e ndlar Africa s lia tia l ai Ch Lebanon Kor Dominican nglaB ru kista xic eece ru zech lta osB mRo itz ni le anym Pe Pa Trinidad ilaza lgium la ger Me Albania Gr Bulga Egypt, C strau Ma Hungary Slovak Slovenia Turke Sw S Sw Spa South weden iwla Minimum Moldova Ma Lithuania Isra Ger Georgia Denmar Croa B Be Be A Al azir 0 Sample size: 56 countries 52 Figure 3. Minimum Balance to Open a Savings Account (% of GDPPC) 80 )CPPDG roonemaC of (% 60 nda ccountA gaU Leone Sierra nyaeK vings Sa 40 orgiaeG Open non to ba riae a racs Le an ga Nig Gh da Ma wila biquema ai s 20 Ma Moz Balance Armen go Moldova Philippine Toba nia a balA an govinaerzeH rd nka s ai . Ethiopia Jo La ry n nia Republic iab India Sri Indone kista fricA eshadlg publiceRk ai publiceRna ndad nda o s Rep y p.eRbarA nia en man xic ni arkm Minimum ungaH Pa Lithua Czech Greece Colom Croatia South Ban stralia Bulgaria Bolivia Slova Ro Malta nicimoD ru Me Pe Swaziland Trinida Brazil iansoB Belaru Slove S edw rea,Ko GermanTurkeySwitzerland Spa Israel Egypt, Den Chile Belgium Au 0 Sample size: 55 countries Figure 4. Annual Fees for a Checking Account (% of GDPPC) Leone Sierra andagU 25 Malawi GDPPC) 20 of (% 15 Account nyaeK 10 Checking Cameroon Swaziland a car Fee an Gh Madagas 5 ai Annual ilehC fricaA Republic Rep. Tobago rzegovinaeH Indones outhS Lebanon rabA and ai and Republic y Republic p.eR Peru liviaoB azilrB lombiaoC LankairS minicanoD ani ian arkm eria stani Moldova Mexico Egypt, Roma Armen Trinidad Bos orgiaeG rusal Turkey Czech German banialA ainSp ovaklS ovenialS ungaryH lgariau eria andr Australia B Switzerland C Korea, oatiar Alg Den B lgiume rael Nig Is Greece Lithuania Sweden ilippineshP Pak Malta Jo India Ethiopia Be ngladeshaB 0 Sample size: 55 countries 53 Figure 5. Annual Fees for a Savings Account (% of GDPPC) 4 i Malaw Uganda GDPPC) 3 of (% Kenya Account 2 Lebanon Bolivia Savings meroonaC nd ilaza Africa Fee Sw 1 South ai iab Herzegovina Indones and Annual Ghana Colom Peru nia Chile Albania osB Georgia Mozambique ainam p.eR Tobago y lia Arab publiceR publiceR and one car Ro Le p.eR licbup h k s a Mexico ades India ai n tia iar ai Turke strauA ea, a ru Egypt, ilz ndlare eece itz Lanka rra lta le anym Rehce lgium la Spain ger rd meni ger Bra Gr Slovak Trinidad Sw Sweden Sri Slovenia Sie Philippines Pakistan Ni Moldova Ma Madagas Lithuania Kor Jo Isra Hungary Ger Ethiopia Dominican Denmar Cz Croa Bulga Be Be nglaB Ar Al 0 Sample size: 56 countries Figure 6. Number of Documents Needed to Open a Checking Account 5) Tobago of 5 ngladeshaB and (Out Chile Trinidad Leone erraiS andagU Cameroon 4 nyaeK Ethiopia aire a frica Account an A Nig Malawi Gh outhS ai Turkey Indones ilippineshP car Malta aire Republic Colombia Swaziland Australia dagasa ai Alg M men coi Checking 3 Ar Mex stan India azilrB oD minican i Pak LankairS LebanonBolivia Greeceu Open Per oldova tia M rzegovinaeH to Croa andr p.eR and Jo Korea, 2 ovenialS ian iar Belgium Bos Bulga orgiaeG Republic Needed Lithuania ungaryH ovaklS rusla karm nia Be Den Roma raelIs Republic Switzerland ain o zambique Sweden Sp M CzechbanialA 1 Documents of 0 Number Sample size: 54 countries 54 Figure 7. Number of Documents to Open a Savings Account 5) 5 of Bangladesh (Out one Le rra 4 Sie Tobago Account and Africa Ghana Cameroon Trinidadltaa raliat M South Savings Uganda Swaziland Aus awi 3 Kenya Mal Madagascar Indonesia nta y Open India kis Republic Pa Turke Lebanon eece co to Bolivia Gr Colombia Philippines azil Armenia Mexi Br Ethiopia and Moldova Jor Croatia Nigeria Dominican ru 2 Slovenia Pe Republic Herzegovina Belgium Georgia Bulgaria and Needed Chile Slovak Rep. Bosnia ea, Denmark Republic Kor su Switzerland edenwS Lanka in Sri Spa Mozambique Lithuania Hungary Czech Belar Algeria Albania 1 Documents of Number 0 Sample size: 52 countries Figure 8. Locations to Submit a Loan Application a c go Afric n eece liarat publi l Rena obaT Spai South Gr Denmark Chile Aus 5 azirB nici ndad Dom inidarT non baeL le Isra nia ithuaL lta eykr 5) Ma Mexico Tu ce eR .p an public of Fr 4 ea,r Rek Ko iab ci Slova tia iar (Out ry nd Colom Croa Bulga nya Republ lare p. sia govinaerz oni Hunga Ke Peru Czech itzwS anst Reb He Indone Paki nkaaL we Ara ab nda cati 3 Sri mb Zi E gypt, Bolivia Bosnia s Appl Ghana Moldova orgia lgium ra sc Ge Be India ga Philippine da biquema esh roonem nia Ma Loant Moz Ca Slove Malawi adlgnaB nia nda Jordan hiopiat oneeL Alba Uga E Armenia Algeria 2 rra Sie Submiot oni 1 Locat 0 Sample size: 52 countries 55 Figure 9. Minimum Amount Required for Consumer Loans 400 ilippineshP GDPPC) of 300 (% Loan Malawi banialA andagU 200 nyaeK Ethiopia Leone Consumer andr stani Jo Pak erraiS a an Gh Amount Bolivia 100 Cameroon aire ai rzegovinaeH car Republic Tobago Rep. Alg irS orgiaeG Lanka and and Minimum ldovao zambiqueo ai Republic Lebanon Indones Republic M M ngladesha dagasa India B u ian rabA p.eR Zimbabwe M Per Malta Bos men Colombia Ar Bulgaria minicanoD fricaA Greece Turkey ovaklS ain ico Czech Sp Chile Trinidad rusla karm Mex Australia S outh Lithuania Egypt, ungary Belgium H Korea, Croatia Be azilrB ovenialS Switzerland Den 0 Sample size: 51 countries Figure 10. Minimum Amount Required for SME Loans 10,000 Bangladesh 8,000 GDPPC) of (% 6,000 Loan SME 4,000 Uganda Amount Georgia 2,000 Herzegovina Ghana and Minimum Albania Tobago Lebanon one Cameroon Philippines Ethiopia Armenia Bolivia Bosnia and Le ewb Republic Rep. nta and Republic Arab Jor rra Republic kis Malta Sie Colombia Zimba Pa ocix Rep. y Africa ea, su Kenya oldova ru eece India ozambique Lanka in adagascar raliat Chile Bulgaria Me M Hungary Slovak Pe Dominican Gr M Belgium Croatia Sri Spa Turke Lithuania M Kor South Switzerland Aus Trinidad azil Br Slovenia Czech Denmark Egypt, Belar Algeria 0 Sample size: 49 countries 56 Figure 11. Fees on Consumer Loans 25 ilehC publiceRna inic Dom Amount) 20 Peru Loan 15 Minimum of (% roonemaC niae 10 Arm Loan l a Brazi bia Afric . go ry Turkey 5 olomC Rep govinaerz South Toba Hunga Greece ltaa nia oldova rea, abweb nia p.eRb He Consumer ra M Alba M oliviaB nda Leone Ko Zim ark Ara nda s sc Ghana Lithua Uga erra n nya coi tia h non iar ndad ga n Si Fee Denm Spai Ke Mex oarC liaa nia Egypt, osniaB i es ba Republic Le ulgaB daa orgia da M Austr Ge Philippine Jor Trinida aw Slove adlg nka India Mal Czech La ndlare Ban Sri itzwS Ethiopia lgiumeB Algeria 0 Sample size: 46 countries Figure 12. Fees on SME Loans public Rena nici 30 Dom Amount) 20 Loan liaratsuA Minimum of (% lgium Loan 10 Be eecerG SME Fee roonem go ry govinaerz Ca l we racs a obaT p. ab ga He public . ci Hunga azirB mb da iar kr Reb Zi Ma eykr oneeL Afric ndad rra nma nda non Rek nda epR Ara Bulga Tu nya Sie De wial an nia nia Republ Ke South Uga Moldova Ma Gha baeL ocxi Me inidarT orgia nia iab esh Slova Bos ea,roK n an nia tia rd Spai Jo Alba Ge Croa India gypt,E ithuaL ani Bolivia Czech hiopiatE Slove ltaa M Colom Arme Peru adlgnaB ndlare itzwS ari ge Al 0 Sample size: 44 countries 57 Figure 13. Days to Process a Consumer Loan Application nta kis Pa 20 15 Applications Loan Philippines 10 Albania Ghana Bangladesh ozambique adagascar su M M Consumer Belar Algeria Lanka Rep. Herzegovina Sri Arab and Hungary Ethiopia Egypt, Bosnia coi Process Bolivia Mex Indonesia Bulgaria Cameroon to ancerF 5 Armenia India Chile y Republic Tobago Georgia and Rep. one Republic and Turke Le Days Belgium Jor Kenya Colombia Croatia Lithuania ea, ru rra awi Africa ewb Republic Pe Kor Dominican Slovak Sie Mal Lebanon South Zimba Switzerland oldova ltaa Uganda M M Trinidad in Slovenia Spa leraIs eece raliat azil Gr Czech Br AusDenmark 0 Sample size: 53 countries Figure 14. Days to Process a SME Loan Application h Banglades 40 nastki Pa Philippines ari ge n a Al 30 Gha Applications Mozambique Loan Rep. SME 20 Arab Republic Lebanon Madagascar Ethiopia Albania Egypt, Chile Process Bulgaria Dominican Republic Herzegovina Tobago Leone and to Czech Lanka India Sri F ancer ocxi and Me Bolivia Indonesia Sierra n 10 Cameroon Bosnia anie Lithuania Colombia Days Jorda Hungary Arm TrinidadtraliasuA e Belarus ltaa Africa Republic M Kenya Georgia Rep. Croatia Turkey Uganda Moldova South ru Zimbabw Slovenia Pe Brazil Belgium Slovak itzerlandwS rea,oK eecerG iSpa n kra Israel nm De 0 Sample size: 52 countries 58 Figure 15. Cost to Transfer Funds Internationally (% of US$250) R Dominican 20 250) of Brazil Ghana (% 15 Swaziland ia liv Bo dova ca Mol banoneL Afrih a Internationally 10 Sout ameroonC huanitiL eden a Kenya stralia Sw Rep. Au bani a Leone na Al Funds Greece Korea, Georgi erraiS u a wia n ye ci Per rk Indi Mal Spai Tu Armenia ltaa M arialg publeR Herzegovi obagoT Jordan Bu Republic ovak anda and Transfer 5 Sl Madagascar Denmark Czech snioB addinirT d tiaa a a to Hungary Cro itzerlanwS oveni Sl a p.eR Indonesi Cost Bangladesh opihitE s Arab, Belaru gyptE miu Uganda Belg 0 Sample size: 44 countries Figure 16. Fees for Using ATM Cards (% of US$100) nta kis Pa .6 ai Niger 100) ocix of Me .4 Africa (% rdaC South ATM liviaoB ru Pe leraIs Rep. ea, ai Republic ing Republic Kor Alger Us Czech .2 lombiaoC Ghana Uganda Slovak for iar Kenya Georgia lgauB l Fee azi obagoT Br iwla and menia Ma Ar Herzegovina Rep. inidadrT and lta Ma ian ai Arab k ye ndlare one car Le lia Bos rk in rra tia tra Albania Tu Switz wedenS oldova dagasa Spa Slovenia Sie Philippines M M banoneL andr eece Jo Indones India Gr hiopiatE gypt,E Denmar oarC ilehC Cameroon lgiumeB sura Bel Aus 0 Sample size: 47 countries 59